Research Collection
Doctoral Thesis
Molecular mechanisms controlling lymphatic vascular function
Author(s): Shin, Jae (Jay) Woo
Publication Date: 2008
Permanent Link: https://doi.org/10.3929/ethz-a-005708685
Rights / License: In Copyright - Non-Commercial Use Permitted
This page was generated automatically upon download from the ETH Zurich Research Collection. For more information please consult the Terms of use.
ETH Library DISS. ETH NO. 17752
MOLECULAR MECHANISMS CONTROLLING LYMPHATIC VASCULAR FUNCTION
A dissertation submitted to ETH Zurich
for the degree of Doctor of Sciences
presented by
JAE (JAY) WOO SHIN April 4th, 1981
citizen of United States of America
accepted on the recommendation of Prof. Michael Detmar Prof. Dario Neri
2008
TABLE OF CONTENTS
1 SUMMARY...... 6
1.1 Summary ...... 7
1.2 Zusammenfassung...... 9
2 INTRODUCTION...... 12
2.1 CHARACTERISTICS OF THE LYMPHATIC VASCULATURE...... 13 2.1.1 Anatomy and physiology of the lymphatic vasculature...... 13 2.1.2 Genes and mechanisms in lymphatic development ...... 15 2.1.2.1 Endothelial lineage-specific differentiation ...... 16 2.1.2.2 Major molecular markers of lymphatic endothelium ...... 22 2.1.2.3 Key lymphangiogenic growth factors ...... 25 2.1.3 Pathologies of the lymphatic vasculature...... 30 2.1.3.1 Lymphatic Dysfunction, Lymphedema...... 30 2.1.3.2 Lymphatic vessels in inflammation and the immune response ...... 31 2.1.3.3 Lymphatic involvement in tumor metastasis ...... 32
2.2 MICROARRAY TECHNOLOGY...... 35 2.2.1 Oligonucleotide microarray technology ...... 35 2.2.2 Gene expression data analysis...... 36
3 RESULTS AND DISCUSSION ...... 38
3.1 Prox1 promotes lineage-specific expression of FGF receptor-3 in lymphatic endothelium..39 3.1.1 Introduction ...... 39 3.1.2 Results...... 40 3.1.2.1 Ectopic expression of Prox1 in primary BEC upregulates FGFR-3...... 40 3.1.2.2 Prox1 binds to the FGFR-3 promoter and activates its transcription ...... 43 3.1.2.3 Identification of the putative Prox1 binding sites in the FGFR-3 promoter ...... 45 3.1.2.4 Expression of FGFR-3 in developing lymphatic vessels of mouse embryo and of human skin...... 46 3.1.2.5 Signaling through FGFR-3 promotes LEC proliferation ...... 47 3.1.2.6 FGF-2 binds directly to low and high affinity receptors in LECs and subsequently internalized for degradation...... 48 3.1.2.7 FGF signaling regulates migration, proliferation and apoptosis of cultured primary lymphatic endothelial cells ...... 50 3.1.3 Discussion...... 52
3.2 Quantification of vascular lineage-specific differentiation and molecular characterization of in vivo (lymph)angiogenesis by a novel low-density microvascular differentiation array ...... 56 3.2.1 Introduction ...... 56 3.2.2 Results...... 57 3.2.2.1 Comprehensive identification of vascular lineage-specific gene signatures ...... 57 3.2.2.2 Identification of lineage-specific biological functions by in silico analysis...... 60 3.2.2.3 Using the LD-MDA to quantify lineage-specific endothelial cell differentiation...... 63 3.2.2.4 Hierarchical clustering according to endothelial lineage-specific gene signatures...... 65 3.2.2.5 Identification of (lymph)angiogenic-mediators using a novel Prediction Relevance Ranking analysis ...... 66 3.2.3 Discussion...... 68
3.3 Lymphatic-specific expression of dipeptidyl peptidase IV and its dual role in lymphatic endothelial function...... 72
3 3.3.1 Introduction ...... 72 3.3.2 Results...... 73 3.3.2.1 Enhanced expression of DPPIV/CD26 by LEC as compared to BEC...... 73 3.3.2.2 Lymphatic vessels in normal skin specifically express DPPIV...... 74 3.3.2.3 DPPIV is expressed by lymphatic vessels in several human organs...... 76 3.3.2.4 Diprotin A inhibits the enzymatic activity of DPPIV but does not induce LEC proliferation and migration...... 78 3.3.2.5 SiRNA-mediated knockdown of DPPIV inhibits LEC adhesion, migration and tube- formation 79 3.3.3 Discussion...... 80
3.4 Transcriptional profiling of VEGF-A and VEGF-C target genes in lymphatic endothelium reveals endocan as a novel mediator of lymphangiogenesis...... 83 3.4.1 Introduction ...... 83 3.4.2 Results...... 85 3.4.2.1 Microarray analysis reveals novel mediators of VEGF-A and VEGF-C-induced effects on lymphatic endothelial cells...... 85 3.4.2.2 ESM-1 expression is potently induced in LEC by VEGF-A and VEGF-C ...... 89 3.4.2.3 ESM-1 promotes LEC proliferation and migration induced by VEGF-A and VEGF-C 92 3.4.2.4 ESM-1 promotes lymphatic vessel activation by VEGF-A in vivo ...... 95 3.4.3 Discussion...... 96
4 CONCLUSIONS AND OUTLOOK ...... 100
4.1 Conclusions ...... 101
4.2 Outlook ...... 103
5 MATERIALS AND METHODS ...... 105
5.1 In Vitro ...... 106 5.1.1 Cell culture...... 106 5.1.1.1 Isolation of human dermal BEC and LEC ...... 106 5.1.1.2 Cells...... 106
5.2 Target validations ...... 107 5.2.1 Electrophoretic mobility shift assay...... 107 5.2.1.1 GST-Prox1-DNA complex...... 107 5.2.2 Quantification of RNA ...... 108 5.2.2.1 Detection and quantification of FGF receptor, DPPIV and ESM1 expressions using qRT-PCR ...... 108 5.2.3 Protein...... 109 5.2.3.1 Detection of DPPIV and ESM1 proteins using Western blotting...... 109 5.2.4 In situ expression validations ...... 109 5.2.4.1 Tissue samples ...... 109 5.2.4.2 Immunofluorescence staining of FGFR-3 in human skin and mouse embryo ...... 110 5.2.4.3 Immunofluorescence of psoriatic skin ...... 110 5.2.4.4 Immunostains of DPPIV ...... 111 5.2.5 Cell culture-based (in vitro) assays...... 112 5.2.5.1 Construction of mutant Prox1 and FGFR-3 reporter gene luciferase assays...... 112 5.2.5.2 Binding and internalization of 125I-FGF-2...... 112 5.2.5.3 Cell proliferation, migration, apoptosis assays and functional inhibition of FGFR-3 113 5.2.5.4 LEC and BEC proliferation assays for FGF-12 and IL7...... 114 5.2.5.5 DPPIV enzyme activity assay...... 115 5.2.5.6 LEC transwell migration, scratch-wound, tube formation and adhesion assays and functional inhibition of DPPIV ...... 115
4 5.2.5.7 Functional inhibition of ESM-1, receptor blocking experiment and LEC proliferation and migration assays ...... 117
5.3 In Vivo ...... 118 5.3.1 Mouse experiments ...... 118 5.3.1.1 ESM-1 siRNA Matrigel assay, immunofluorescence stainings and morphometric analyses 118
5.4 TRANSCRIPTOMICS...... 119 5.4.1 Gene expression profiling using oligonucleotide microarrays...... 119 5.4.1.1 Gene expression profiling of human LEC and BEC ...... 119 5.4.1.2 LEC vs. BEC microarray data analysis...... 119 5.4.1.3 Gene expression profiling for VEGF-A and VEGF-C target genes in LEC...... 120 5.4.1.4 VEGF-A, VEGF-C stimulated time course data analysis...... 120 5.4.1.5 Establishment of a Low-Density Microvascular Differentiation Array...... 121 5.4.1.6 Endothelial lineage score analysis and identification of core signature genes ...... 122 5.4.2 Bioinformatics ...... 123 5.4.2.1 Microarray data mining tools...... 123 5.4.2.2 Prediction Relevance Ranking (PRR) analysis...... 123
6 BIBLIOGRAPHY ...... 125
7 APPENDIX...... 150
8 CURRICULUM VITAE ...... 184
9 ACKNOWLEDGEMENTS ...... 187
5 1 SUMMARY
6 1.1 Summary
In this thesis, I describe several novel molecular mechanisms controlling lymphatic vascular function during endothelial lineage-specific differentiation and lymphangiogenesis.
The essential functions of the lymphatic vascular system include the maintenance of tissue fluid homeostasis and immune surveillance. Impaired function of the lymphatic system can lead to several diseases such as primary or secondary lymphedema, whereas recent evidence indicates that tumor-induced activation of lymphatic vessels promotes cancer metastasis. In the past, the lack of lymphatic-specific molecular markers has hampered progress in the field of lymphatic vascular biology. However, during the last decade, several key lymphatic-specific markers have been discovered and have been shown to be important molecular regulators during embryonic development, normal fluid balance homeostasis, the afferent immune response, acute and chronic inflammation and cancer spread. In this thesis, I have investigated novel molecular mechanisms regulating lymphatic vascular function, based on the identification of novel lymphatic-specific markers by oligonucleotide microarrays of cultured lymphatic endothelial cells, and on the functional characterization of select lymphatic-specific markers in vitro and in vivo.
In order to investigate the role of the lymphatic system during embryogenesis, we have recently overexpressed the lymphatic-specific transcription factor Prox1 in cultured blood vascular endothelial cells (BEC), isolated from human foreskin. We found that ectopic overexpression of Prox1 recapitulates, at least in part, the embryonic lymphatic reprogramming of vascular endothelium by downregulating BEC-specific genes and by up-regulating several lymphatic endothelial cell (LEC)- specific genes. In this thesis, I present evidence demonstrating that Prox1 upregulates the expression of Fibroblast Growth Receptor-3 (FGFR-3) during lymphatic reprogramming and that FGF signaling through the upregulated FGFR-3 plays an important role in the early development of the lymphatic vascular system (Chapter 3; Section 1).
7 Furthermore, using transcriptional profiling by gene microarray technology, I have compared the gene expression profiles of cultured human blood vascular (BEC) and lymphatic (LEC) endothelial cells. These studies have revealed a set of 236 lymphatic signature genes and 342 blood vascular signature genes, many of which have not been previously known to be expressed in a lineage-specific manner. Based on the identification of these signature genes, I have established a Low-Density Microvascular Differentiation Array (LD-MDA), a novel tool to quantify the degree of endothelial lineage-specific differentiation of various endothelial cell types in vitro which has also allowed the identification of novel (lymph)angiogenesis factors involved in the chronic inflammatory skin disease psoriasis (Chapter 3; Section 2).
Based on the identification of lymphatic signature genes, I present evidence that an active form of dipeptidyl peptidase IV (DPPIV) is more strongly expressed in lymphatic endothelium as compared to blood vascular endothelium in several different human tissues. To investigate the functional role of DPPIV in LEC biology, I have performed cell proliferation, migration, tube formation and adhesion assays after siRNA-mediated knockdown of DPPIV. These studies have elucidated a dual function of DPPIV in lymphangiogenesis (Chapter 3; Section 3).
Previous studies have revealed that vascular endothelial growth factor-A (VEGF-A) and VEGF-C are upregulated in metastatic cancers, and that they are the major molecular mediators of tumor-induced lymphangiogenesis which promotes lymph node and distant cancer metastasis. Thus, the identification of downstream mediators of the effects of VEGF-A and/or VEGF-C may reveal novel targets for inhibiting lymphangiogenesis and cancer spread. In this thesis, I have performed a comprehensive gene expression profiling screen of LEC stimulated with VEGF-A or VEGF-C for different periods of time. These studies have revealed a number of novel mediators of lymphangiogenesis and, in particular, have identified endocan - also known as ESM1 - as a novel mediator of VEGF-A and of VEGF-C-induced lymphangiogenesis. I demonstrate that endocan significantly promotes LEC proliferation and migration in concert with VEGF-A and VEGF-C, and that silencing of endocan expression significantly attenuates the VEGF-A/VEGF-C induced LEC proliferation and migration in vitro and VEGF-A induced lymphatic vessel enlargement in vivo (Chapter 3; Section 4).
8 1.2 Zusammenfassung
In dieser Arbeit beschreibe ich mehrere neue molekulare Mechanismen, welche die Differenzierung lymphatischer Endothelzellen sowie die Lymphangiogenese kontrollieren.
Zu den essentiellen Funktionen des lymphatischen Systems gehören die Regulierung des Flüssigkeitsdrucks im Gewebe sowie die Immunüberwachung des Organismus. Eine Funktionsstörung des lymphatischen Systems kann zu einer Reihe von Krankheiten führen, zum Beispiel dem primären oder sekundären Lymphödem. Jüngste Studien weisen zudem darauf hin, dass die Krebsmetastasierung durch eine Tumor-induzierte Aktivierung lymphatischer Gefässe gefördert wird. Das Fehlen von spezifischen molekularen Markern für lymphatische Gefässe hat die Erforschung der Biologie des lymphatischen Systems lange Zeit behindert. Im vergangenen Jahrzehnt wurden jedoch mehrere entscheidende solcher Marker entdeckt, welche sich auch als wichtige molekulare Regulatoren erwiesen haben für die embryonale Entwicklung des lymphatischen Systems, die Regulierung des Gewebedrucks und die afferente Immunantwort, sowie für pathologische Situationen wie die akute und chronische Entzündung und die Ausbreitung von Krebs. Basierend auf der Identifizierung neuer spezifischer lympatischer Marker mittels Oligonukleotid-Microarrays kultivierter lymphatischer Endothelzellen, sowie der funktionellen Charakterisierung ausgewählter solcher Marker in vitro und in vivo, habe ich in dieser Arbeit neue molekulare Mechanismen untersucht, welche lymphvaskuläre Funktionen regulieren.
Um die Rolle des lymphatischen Systems während der Embryogenese zu untersuchen, haben wir kürzlich den Lymphendothel-spezifischen Transkriptionsfaktor Prox1 in aus menschlicher Vorhaut isolierten, kultivierten Blutgefässendothelzellen überexprimiert. Es zeigte sich, dass die Überexpression von Prox1 zumindest teilweise die Umprogrammierung von Blutgefäss- zu Lymphendothelzellen während der embryonalen Entwicklung rekapituliert, indem sie die Expression Blutgefässendothel-spezifischer Gene verringert und die Expression mehrerer Lymphendothel-spezifischer Gene erhöht. Meine Arbeit liefert Hinweise darauf, dass Prox1 während der lymphatischen Umprogrammierung des Blutgefässendothels die
9 Expression von FGFR-3 verstärkt, und dass FGF-Signale, welche über diesen verstärkt exprimierten Rezeptor vermittelt werden, während der frühen Entwicklung des lymphatischen Systems eine wichtige Rolle spielen (Kapitel 3; Abschnitt 1).
Des weiteren habe ich die Genexpressionsprofile kultivierter humaner Blutgefäss- und Lymphendothelzellen mittels Microarray-Technologie verglichen. Diese Untersuchungen führten zu einem Satz von 236 lymphatischen Signatur-Genen und 342 Blutgefäss-Signatur-Genen, von denen viele noch nicht als spezifisch für den einen oder anderen Endothelzelltyp bekannt waren. Basierend auf der Identifizierung dieser Signatur-Gene habe ich sogenannte „Low Density Microvascular Differentiation Assays“ (LD-MDA) entwickelt, mit deren Hilfe sich das Ausmass der lymphatischen oder blutgefässartigen Differenzierung verschiedenster Endothelzell- Typen in vitro quantifizieren lässt. Diese Assays haben auch die Identifizierung neuer (lymph)angiogener Faktoren ermöglicht, welche in der chronisch entzündlichen Hautkrankheit Schuppenflechte (Psoriasis) eine Rolle spielen (Kapitel 3; Abschnitt 2).
Ebenfalls basierend auf der Identifizierung lymphatischer Signatur-Gene zeige ich, dass eine aktive Form des Enzyms Dipeptidylpeptidase IV (DPPIV) stärker auf lymphatischem als auf Blutgefässendothel exprimiert wird. Um die funktionelle Bedeutung von DPPIV in lymphatischen Endothelzellen zu untersuchen, habe ich Zellproliferations-, Migrations-, Röhrenbildungs- und Adhäsions-Analysen nach Ausschaltung der DPPIV-Expression mittels siRNA durchgeführt. Diese Versuche haben eine Doppelfunktion von DPPIV in der Lymphangiogenese gezeigt (Kapitel 3; Abschnitt 3).
Aus früheren Studien ist bekannt, dass die Wachstumsfaktoren Vascular Endothelial Growth Factor-A (VEGF-A) und VEGF-C in metastatischen Krebsgeschwüren hochreguliert sind und dass sie die bedeutendsten molekularen Mediatoren der Tumor-induzierten Lymphangiogenese sind, welche ihrerseits Lymphknoten- und entfernte Metastasen begünstigt. Die Identifizierung molekularer Mediatoren der Effekte von VEGF-A und/oder VEGF-C könnte daher zu neuen Zielmolekülen für die Hemmung der Lymphangiogenese und damit der Krebsausbreitung führen. In dieser Arbeit habe ich eine umfassende Untersuchung der Genexpressionsprofile lymphatischer Endothelzellen, welche während unterschiedlicher Zeiträume mit
10 VEGF-A oder VEGF-C stimuliert wurden, durchgeführt. Diese Studien haben eine Reihe neuer Mediatoren der Lymphangiogenese ergeben, insbesondere habe ich Endocan – auch bekannt als ESM1 – als neuen Mediator der durch VEGF-A und durch VEGF-C induzierten Lymphangiogenese identifiziert. Ich zeige, dass Endocan im Zusammenspiel mit VEGF-A und VEGF-C die Proliferation und Migration lymphatischer Endothelzellen signifikant fördert, und dass Ausschaltung der Expression von Endocan die durch VEGF-A/C induzierte Proliferation und Migration lymphatischer Endothelzellen in vitro sowie die durch VEGF-A induzierte Vergrösserung lymphatischer Gefässe in vivo signifikant abschwächt (Kapitel 3; Abschnitt 4).
11 2 INTRODUCTION
12 2.1 CHARACTERISTICS OF THE LYMPHATIC VASCULATURE
2.1.1 Anatomy and physiology of the lymphatic vasculature
The lymphatic system is composed of a vascular network of thin-walled lymphatic capillaries and thick-walled collecting lymphatic vessels. Unlike blood vessel capillaries, the lymphatic capillaries consist of a single-cell layer of overlapping, flat endothelial cells and are blind-ended structures lacking pericytes, smooth muscle cells and a basement membrane. Endothelial cells in lymphatic capillaries form loose intercellular valve-like junctions and exhibit large interendothelial pores. Anchoring filaments consisting of microfibrils and elastin connect the lymphatic endothelial cells to the extracellular matrix. One of the main functions of the lymphatic vasculature is the maintenance of fluid homeostasis by absorbing interstitial fluid, or lymph. Macromolecules and cells, including extravasated leukocytes, leaked from blood capillaries, as well as activated antigen-presenting cells are taken up by lymphatic capillaries. From here, lymph is transported towards collecting lymphatic vessels (Oliver & Detmar, 2002).
The collecting lymphatic vessels have a smooth muscle cell layer, basement membrane and valves to prevent back flow. The contraction of smooth muscle cells and surrounding skeletal muscles, as well as arterial pulsations, contribute to lymph propulsion (Leu et al, 1999; Petrova et al, 2004). These larger vessels drain into either one of two collecting vessels. The main, longer trunk is the thoracic duct, which runs parallel with the aorta. The thoracic duct empties lymph into the blood stream at the junction of the left subclavian vein with the left internal jugular vein located at the base of the neck. Another, shorter collecting trunk is the right lymphatic duct, which empties its lymph into the right subclavian vein (Ambrose, 2006; Hong et al, 2004c) (Fig. 2.1.1).
13 Heart
Artery
~ Lymphalic vessel
Peripheral tissue capillaries
Figure 2.1.1 Schematic illustration of the blood vascular and lymphatic system. The blood vascular system is a circular and closed system, whereas the lymphatic system is open-ended and linear. Fluids, macromolecules, and cells extravasated from blood capillaries flow into lymphatic capillaries in peripheral tissues and are then transported by means of the larger collecting lymphatic vessels and the thoracic duct back to the blood vascular system for recirculation (Hong et al, 2004c).
Another important function of the lymphatic system is immune surveillance (Massberg et al, 2007). The lymphatic system includes lymphoid organs such as the lymph nodes, tonsils, Peyer’s patches, spleen, and thymus, all of which play an important role in the immune response (Heydtmann et al, 2006; Millington et al, 2007). Lymph, which contains memory T cells, antigens, antigen-bearing dendritic cells and macrophages, is normally filtered through the lymph nodes through collecting-type terminal afferent lymphatics in local peripheral tissues (e.g. skin). It then percolates through the lymphoid tissue and a series of sinuses, and exits the node in the efferent lymphatics (Daynes et al, 1985). The cellular component of the lymph node includes T-cell-dependent paracortical areas, in which naïve T cells from neighboring venules are brought into contact with antigen-presenting dendritic cells (Cavanagh & Von Andrian, 2002). B cells are mainly associated with the germinal follicles in the outer cortex, where naïve B cells acquire the capacity to synthesize epitope-specific antibodies (Randolph et al, 2005). Lymphatic vessels are not
14 normally present in avascular structures such as the epidermis, hair, nails, cartilage, and cornea, nor are they present in some vascularized organs such as brain and retina (Niederkorn et al, 1989; Oliver & Detmar, 2002).
2.1.2 Genes and mechanisms in lymphatic development
The earliest description of the lymphatic system dates back to the 5th century B.C. Hippocrates’ work entitled On Joints stated that “all men have glands, smaller or larger, in the armpit and many other parts of the body” describing the lymph nodes (Withington, 1894). However, the understanding of the lymphatic system only started to augment in the 17th century when Gasparo Aselli (1581 – 1626), a physician in Milan and later Professor of Anatomy in Pavia, observed the lacteals – lymphatic vessels in the intestine - while dissecting a living, well-fed dog, something he had never seen in fasting dogs (Gasparo, 1627). Originally described as “milky veins”, the mechanisms controlling the normal development of lymphatic vessels and the molecular regulation of their biological function have remained unclear until the early 20th century. Florence Sabin proposed in 1902 that, in vertebrates, the endothelial cells bud off from the veins during early embryonic development and form primitive lymph sacs. The peripheral lymphatic system then originates from these primary lymph sacs by endothelial sprouting into the surrounding tissues and organs, where local capillaries are formed (Sabin, 1902). A few years after Sabin’s “centrifugal” proposal, Hungtington and McClure proposed that the first lymphatics arise independently in the mesenchyme and that they are connected to the venous system only later (Huntington & McClure, 1910). The controversy over the origin of lymphatic vasculature was not resolved until recently. Wigle and Oliver studied mice deficient in the homeodomain protein Prox1 and found that these mice were unable to develop a lymphatic vascular system and that Prox1 was required for a subset of venous endothelial cells in the embryonic cardinal veins to migrate out and to form the initial lymphatic vessels during early embryogenesis (Wigle et al, 2002a). These studies also identified some of the molecular determinants that control the step-wise process of lymphatic competence, commitment, differentiation and maturation (Oliver & Harvey, 2002) as described in the following section:
15 2.1.2.1 Endothelial lineage-specific differentiation
During embryogenesis, an unknown molecular factor(s) regulates the initial stage of lymphatic competence by inducing the lymphatic vessel endothelial hyaluronan receptor 1 (Lyve1) (Banerji et al, 1999) on a few of the endothelial cells that line the anterior cardinal vein of mice at embryonic day (E) 8.5-9.5. This could be considered the first morphologic indication that venous endothelial cells are already competent to respond to a lymphatic-inducing signal (Oliver, 2004). A few hours following the expression of Lyve1 by venous endothelial cells in mice, expression of the transcription factor Prox1 is observed in a subpopulation of venous endothelial cells in a polarized manner. Subsequently, these lymphatic endothelial cell progenitors bud off, proliferate and migrate to form the embryonic lymph sacs and lymphatic vascular network. These cells also express the receptor tyrosine kinase vascular endothelial growth factor receptor-3 (VEGFR-3 or FLT4), the cell surface receptor that binds the lymphangiogenic growth factors, VEGF-C and VEGF-D (Makinen et al, 2005). By E14.5 – 15.5, the primary jugular lymph sacs subsequently sprout to form a primitive lymphatic plexus, which spreads throughout the head and neck, thorax and forelimbs (Oliver & Harvey, 2002). Further maturation and differentiation of the lymphatic vasculature occurs in a progressive manner until the first postnatal days (Fig. 2.1.2) (Karpanen et al, 2006; Oliver, 2004; Saharinen et al, 2004).
16 F-
Lymphatic Lymphatic Lymphatic specification competence commitment budding and migration
Figure 2.1.2 Current model of the stepwise embryonic development of the mammalian lymphatic system. At mouse embryonic day (E) 8.5, all endothelial cells of the cardinal vein express the lymphatic markers LYVE-1 and vascular endothelial growth factor recpetor-3 (VEGFR-3) and display lymphatic competency. Upon stimulation by a yet unidentified inductive signal, a subset of venous endothelial cells becomes lymphatically biased and up-regulates Prox1 around E10.5. At E11.5 and thereafter, these Prox1-positive cells bud off and migrate out to form initial lymphatics. They also up- regulate the expression of additional lymphatic-specific molecules such as podoplanin and secondary lymphoid chemokines (SLC). The formation of a mature lymphatic network continues through the first postnatal days.
In the following section, a detailed description of genes and mechanisms that have been identified as mediators for lymphatic endothelial lineage-specific differentiation is outlines:
2.1.2.1.1 Prox1
Prox1 is a homeodomain protein that was originally isolated due to its high homology to the Drosophila protein prospero (Oliver et al, 1993a; Tomarev et al, 1996). Prox1 is an important regulator of cell differentiation and embryogenesis in several tissues such as developing liver, nervous system, pancreas, lens, retina, heart and lymphatic vessels (Oliver et al, 1993a; Sosa-Pineda et al, 2000; Wigle et al, 2002a). Thus far, Prox1 proteins have been identified in human, mouse, chicken, newt, frog, and zebrafish; and their amino acid sequences are highly conserved across these species (Oliver et al, 1993a; Tomarev et al, 1996). In mice, Prox1 expressing endothelial cells are first observed at E10.5 in the jugular vein, from which they migrate to form the first lymphatic sprouts. Endothelial cells of Prox1-/- mice bud from the cardinal vein but fail to express lymphatic endothelial markers and do not migrate further (Wigle et al, 2002a). These findings suggested that Prox1 might specify lymphatic cell fate by
17 directly reprogramming the transcriptome of embryonic venous endothelial cells. Indeed, Prox1 overexpression in human blood vascular endothelial cells (BEC) suppresses many blood vascular-specific genes and upregulates lymphatic endothelial cell (LEC)-specific transcripts (Hong et al, 2002; Petrova et al, 2002a). Prox1 is also shown to be upregulated in cultured BEC when infected with Kaposi’s sarcoma- associated herpes virus (KSHV, also known as HHV-8). This KSHV-mediated upregulation of Prox1 leads to reprogramming of BEC to adopt LEC phenotypes by inducing the expression of more than 70% of major lymphatic-associated genes and by downregulating many BEC-specific genes (Hong et al, 2004a). This Prox1 mediated cell fate reprogramming in KSHV-infected cells provides an additional support to the concept that Prox1 is a master control gene specifying lymphatic endothelial cell fates. Furthermore, Prox1+/- mice develop chylous ascites, and show disorganized and abnormally patterned lymphatic vessels (Harvey et al, 2005). Notably, impaired lymphatic vascular function in Prox1 heterozygotes and in mice with conditional deletion of Prox1 in endothelial cells causes adult onset obesity, indicating an important link between lymphatic function and adipogenesis. So far, treatment of cultured endothelial cells with interleukin-3 and -7 was shown to induce the expression of Prox1; however, the relevance for the in vivo regulation of Prox1 expression remains unclear (Al-Rawi et al, 2005; Groger et al, 2004).
2.1.2.1.2 VEGFR-3
VEGFR-3, also known as FLT4, is a member of the fms-like tyrosine kinase family and is structurally related to the two VEGF-A receptors VEGFR-1/FLT1 and VEGFR-2/KDR/FLK1 (Kaipainen et al, 1995; Kaipainen et al, 1993). VEGFR-3 was the first gene to be identified as lymphatic vessel-specific (Kaipainen et al, 1995). VEGFR-3 does not interact with VEGF-A but acts as a signaling receptor for VEGF- C and VEGF-D – the two most potent lymphangiogenic factors known so far. VEGFR-3 deletion in mice leads to defects in blood vessel remodeling and embryonic death at mid-gestation, indicating an early blood vascular function (Dumont et al, 1998). During embryonic development, VEGFR-3 is expressed by venous endothelial cells and also by angioblasts of the head mesenchyme during E8.5 to E12.5 (Kaipainen et al, 1995). Only later in embryogenesis, VEGFR-3 expression becomes
18 specific for the lymphatic endothelial cells and is gradually down-regulated by venous endothelial cells. VEGFR-3 is known to be specific for lymphatic vessels in adult tissues, however, some tumor-associated and wound-associated blood vessels re- express VEGFR-3 (Hirakawa et al, 2007; Kubo et al, 2000)
2.1.2.1.3 VEGF-C/D
VEGF-C and VEGF-D were originally cloned as ligands for VEGFR-3 (Achen et al, 1998; Joukov et al, 1996; Marconcini et al, 1999; Orlandini et al, 1996; Yamada et al, 1997). VEGF-C is expressed by a multitude of cell types, including mesenchymal cells around embryonic veins, activated macrophages, skeletal muscle cells, and smooth muscle cells surrounding large arteries (Eichmann et al, 1998; Joukov et al, 1996; Karkkainen et al, 2004; Kukk et al, 1996). Both VEGF-C and VEGF-D are produced as precursor proteins with N- and C-terminal propeptides flanking the VEGF homology domain (Joukov et al, 1997; Stacker et al, 1999). The secreted factors undergo proteolytic processing that results in increased affinity for both VEGFR-2 and VEGFR-3, and consequently increases their ability to induce angiogenesis (Cao et al, 1998) and lymphangiogenesis in vivo (Enholm et al, 2001; Karkkainen et al, 2004; Saaristo et al, 2002). Activation of VEGFR-3 by VEGF-C and/or VEGF-D promotes proliferation, migration, and survival of cultured human LEC (Makinen et al, 2001) and they can also induce lymphangiogenesis in adult tissues (Jeltsch et al, 1997; Veikkola et al, 2001).
The recent inactivation of the Vegf-c gene in mice has provided additional information regarding its role during embryonic lymphangiogenesis (Karkkainen et al, 2004). The mutant embryos showed that Vegf-c activity is essential during the lymphatic development since its functional inactivation results in embryonic lethality. Further analysis demonstrated that Vegf-c activity is essential for promoting the budding and proliferation of Prox1-expressing lymphatic endothelial cells located in the embryonic veins. This suggests that Vegf-c is an essential chemotactic and survival factor during embryonic lymphangiogenesis (Karkkainen et al, 2004). Contrarily, Vegf-d-deficient mice do not exhibit a lymphatic phenotype probably because Vegf-d is not expressed at the critical sites of lymph sac formation in the
19 embryo (Avantaggiato et al, 1998; Karkkainen et al, 2004). However, exogenous VEGF-D protein rescues the impaired vessel sprouting in Vegfc-/- embryos (Karkkainen et al, 2004).
2.1.2.1.4 LYVE-1
LYVE-1 has been identified as a lymphatic endothelium-specific hyaluronan (HA) receptor. HA is a large mucopolysaccharide polymer (105-7 Daltons) that represents a major component of the extracellular matrix in many tissues (Jackson et al, 2001). LYVE-1 is one of the most widely used markers for lymphatic endothelial cells (Jackson, 2004) and in mice, it is the first marker of lymphatic endothelial commitment (Oliver, 2004). In adults, LYVE-1 expression is downregulated in the collecting lymphatic vessels but remains high in lymphatic capillaries (Makinen et al, 2005). However, Lyve-1-deficient mice appear normal and no obvious lymphatic vascular malfunctions or morphological abnormalities have been detected thus far (Gale et al, 2007).
2.1.2.1.5 Syk and SLP76
The tyrosine kinase Syk and the adaptor protein SLP-76 are involved in controlling the separation of the lymphatic and blood vascular systems during embryogenesis. Recent studies indicate that Syk- and SLP-76-mediated hematopoietic signaling might be required to separate emerging lymphatic vessels from the blood vascular system (Abtahian et al, 2003). Syk is widely expressed in hematopoietic cells and is involved in coupling activated immunoreceptors to downstream signaling events that mediate diverse cellular responses, including proliferation, differentiation, and phagocytosis (Yanagi et al, 2001). The adapter protein SLP-76 (also known as lymphocyte cytosolic protein 2, LCP2) is a substrate of Syk for downstream signaling (Clements, 2003). Syk- or Slp-76-deficient mice that survived to adulthood display arterial- venous-lymphatic shunting and, as a result, exhibit cardiomegaly, elevated cardiac output, and admixture of blood with lymph, suggesting that hematopoietic cells might be involved in the separation of the two vascular systems. Additionally, the lymph-
20 vascular phenotype of Slp-76-deficient mice was ameliorated when Slp-76 null bone marrow cells were introduced into lethally irradiated wild-type mice (Abtahian et al, 2003). These findings suggest the possibility that hematopoietic precursor cells might influence the development of the lymphatic system. More recent study indicate that deletion of Spred-1 and Spred-2 resulted in embryonic lethality at E12.5 to 15.5 with marked subcutaneous hemorrhage, edema, and dilated lymphatic vessels filled with erythrocytes (Taniguchi et al, 2007), resembling that of Syk-/- and SLP-76-/- mice with defects in the separation of lymphatic vessels from blood vessels.
Table 2.1-1 Genes that mediate lymphatic vasculature formation and patterning Gene Model Phenotype References Adhesion molecules Integrin α9 KO Respiratory failure caused by pleural fluid (Huang et al., 2000) (chylothorax), lymphedema
Growth factors/receptors Angiopoietin-1 TG Hyperplastic lymphatic vessels (Tammela et al., 2005) Angiopoietin-2 KO Hypoplasia, chylous ascites (Gale et al., 2002) VEGF-C KO No lymphatic vessels (-/-), hypoplasia, chylous (Karkkainen et al., 2004) ascites, lymphedema (+/-) VEGF-C TG Hyperplastic lymphatic vessels (Jeltsch et al., 1997) VEGFR-3 KO Hyperplastic lymphatic vessels; cardiovascular (Dumont et al., 1998) failure VEGFR-3 Chy Lymphedema (Karkkainen et al., 2001) mice Neuropilin-2 KO Lymphedema, reduction of small lymphatic (Yuan et al., 2002) vessels during development HGF TG Enhanced formation and enlargment of (Kajiya et al., 2005) lymphatic vessels
Transcription factors Prox1 KO No lymphatic vasculature developed (-/-), (Wigle and Oliver 1999; Harvey et adult-onset obesity, chylous ascites (+/-) al., 2005) FOXC2 KO Abnormal lymphatic patterning, absent valves, (Kriederman et al., 2003; Petrova et lymphatic dysfunction (-/-), lymphatic vessel al., 2004) and lymph node hyperplasia (+/-) Net (Elk3) KO Chylothorax, dilated lymphatic vessels (Ayadi et al., 2001) SOX18 (ragged) KO Generalized edema and chyle in the (Pennisi et al., 2000) peritoneum
Miscellaneous Podoplanin KO Lymphedema, dilation of lymphatic vessels (Schacht et al., 2003) and diminished lymphatic transport SLP-79 and Syk KO Abnormal blood-lymphatic connections (Abtahian et al., 2003) Ephrin B2 Mutant Defective remodeling of lymphatic vascular (Makinen et al., 2005) network, hyperplasia, lack of valves, chylothorax Adrenomodullin KO Intestitial lymphedema, abnormal lymphatic (Fritz-Six et al., 2008) patterning FIAF KO Dialated intestinal lymphatic vessels (Backhed et al., 2007)
FOXC2, forkhead box C2; HGF, hepatocyte growth factor; KO, knock-out; SLP, Src homology 2-domain containing leukocyte protein; SOX18, sex determining region Y-related high mobility group box 18; TG, transgenic; VEGF, vascular endothelial growth factor; VEGFR, vascular endothelial growth factor receptor; FIAF, fasting-induced adipose factor.
21
2.1.2.2 Major molecular markers of lymphatic endothelium
Major advances in lymphatic research have been made possible by the recent establishment of defined cultures of blood vascular endothelial cells (BECs) and lymphatic endothelial cells (LECs) isolated from human skin (Hirakawa et al, 2003; Kriehuber et al, 2001; Makinen et al, 2001; Podgrabinska et al, 2002). Comparative microarray analyses of their specific transcriptomes revealed that approximately 2% of transcribed genes are differentially expressed between BECs and LECs, and this difference may reflect their distinct in vivo functions (Hirakawa et al, 2003; Petrova et al, 2002a). However, the arrays used in these analyses included an incomplete set of human genes and a large-scale confirmation of the results by other methods has not been attempted. Very recently, surface-accessible proteins of BECs and LECs were biotinylated, purified by high performance liquid chromatography (HPLC) and analyzed by mass spectrometry (Roesli et al, 2008). This technology has provided the first insight into the surface-accessible, vascular lineage-specific proteome.
The following sections describe well-characterized lymphatic vessel markers and their role in lymphangiogenesis:
2.1.2.2.1 Podoplanin
Podoplanin is a mucin-type transmembrane glycoprotein that is highly expressed by podocytes, keratinocytes, cells of the choroid plexus, alveolar type II lung cells, lymphatic endothelial cells, but not by blood vascular endothelial cells (Hirakawa et al, 2003; Kriehuber et al, 2001; Petrova et al, 2002a; Schacht et al, 2003; Wetterwald et al, 1996). During mouse embryonic development, podoplanin is expressed at around E9 in the central nervous system and the foregut but not yet in the vascular system (Rishi et al, 1995; Schacht et al, 2003). At E11.5-E12.5, podoplanin is expressed by all endothelial cells in the cardinal vein, including the budding Prox1 positive cells, but the expression is progressively down-regulated by venous endothelial cells. At birth, the expression of podoplanin is restricted to LEC (Schacht et al, 2003). Podoplanin deficiency leads to abnormal lung development and perinatal
22 lethality. Neonatal podoplanin knockout mice displayed abnormal lymphatic function and patterning, mimicking lymphedema, possibly due to impaired migration and adhesion of lymphatic endothelial cells (Schacht et al, 2003).
2.1.2.2.2 FOXC2
The forkhead transcription factor FOXC2 is involved in the specification of the lymphatic capillary versus collecting lymphatic vessel phenotype. FOXC2 is highly expressed in the developing lymphatic vessels as well as in lymphatic valves in adults (Petrova et al, 2004). FOXC2 is essential for the morphogenesis of lymphatic valves and the establishment of a pericyte-free lymphatic capillary network. Mice heterozygous for Foxc2 exhibited a generalized lymphatic vessel and lymph node hyperplasia and rarely exhibited hindlimb swelling, mimicking closely the distinctive lymphatic and ocular phenotype of lymphedema-distichiasis (LD) patients (Kriederman et al, 2003; Yuan et al, 2002).
2.1.2.2.3 Neuropilin2
Neuropilin-1 (Nrp1) and neuropilin-2 (Nrp2) are transmembrane glycoproteins with large extracellular domains that interact with both class 3 semaphorins and VEGFs. While Nrp1 is predominantly expressed in arterial endothelial cells, Nrp2 is mainly expressed in veins and in visceral lymphatic vessels and weakly expressed in the cutaneous lymphatics (Karkkainen et al, 2001). Homozygous Nrp2 mutants show absence or severe reduction of small lymphatic vessels and capillaries during development. Arteries, veins and larger collecting lymphatic vessels developed normally in these mice, suggesting that Nrp2 is selectively required for the formation of small lymphatic vessels and capillaries (Yuan et al, 2002).
2.1.2.2.4 CCL21
CCL21 (CC chemokine ligand 21), also known as secondary lymphoid chemokine (SLC), plays an important role in immunoregulatory and inflammatory processes.
23 CCL21 is expressed by lymphatic endothelium and is secreted as a 12 KDa protein but it is immobilized by the extracellular matrix (ECM) by binding to sulfated proteoglycans (Patel et al, 2001). It is also expressed in the high endothelial venules and the T cell areas of lymph nodes and Peyer's patches. CCL21 promotes adhesion and stimulates migration of thymocytes, T-lymphocytes, macrophages, and neutrophils through high-affinity binding to chemokine receptor 7 (CCR7) (Gunn et al, 1998; Tangemann et al, 1998). Increased incidence of lymph node metastases has been correlated with the presence of CCR7 on human carcinoma cells (Cabioglu et al, 2005; Gunther et al, 2005; Heresi et al, 2005; Wiley et al, 2001), possibly sensing chemotactic gradients of CCL21 originating from lymphatics (Shields et al, 2007).
2.1.2.2.5 Adrenomedullin
Adrenomedullin (AM) is a multifunctional peptide vasodilator that tranduces its effects through the calcitonin receptor-like receptor (calcrl) when the receptor is associated with a receptor activity-modifying protein (RAMP2) (McLatchie et al, 1998). AM-activated ERK signaling was reported to be greater in human LEC as compared to human BEC and loss of AM signaling resulted in abnormal jugular lymphatic vessels due to reduction in lymphatic endothelial cell proliferation. Additionally, AM-null mice developed intestinal lymphedema and died during mid- gestation (Fritz-Six et al, 2008).
2.1.2.2.6 FIAF/ANGPTL-4
Fasting-induced adipose factor (Fiaf), also known as angiopoietin-like protein 4 (Angptl4), is a glycosylated, secreted, and proteolytically processed protein (Kersten et al, 2000; Kim et al, 2000). Fiaf has been reported to promote endothelial cell survival in the gut after damage from ionizing radiation and reduces VEGF-induced microvascular permeability in the skin (Crawford & Gordon, 2005). Recently, Prox1 has been identified as a downstream target for Fiaf signaling in the intestinal lymphatic endothelium, and Fiaf-deficient mice die during the suckling period with dilated intestinal lymphatic vessels. Fiaf also has been identified as an organ-specific mediator of lymphangiogenesis that is instrumental in sustaining separate blood and
24 lymphatic circulatory systems in the intestine, and in its supporting mesentery after birth (Backhed et al, 2007).
Table 2.1-2 Specific markers for lymphatic vessels versus blood vessels
Markers Function LV BV Reference Prox1 Transcription factor ++ - (Wigle and Oliver, 1999) Podoplanin Transmembrane glycoprotein ++ - (Wetterwald et al., 1996; Breiteneder- Geleff et al., 1999) LYVE-1 Hyaluronan receptor ++ - (Banerji et al., 1999) VEGFR-3 Growth factor receptor + - (Kaipainen et al., 1995) /(+)1 Neuropilin-2 Semaphorin and growth factor receptor + - (Yuan et al., 2002) /(+)2 Macrophage L-selectin receptor + - (Irjala et al., 2001) mannose receptor 1 CCL21 CC-chemokine + - (Gunn et al., 1998) Desmoplakin Anchoring protein of adhering junctions + - (Ebata et al., 2001) Integrin α9 Adhesion molecule, subunit of + - (Huang et al., 2000; Petrova et al., osteopontin and tenascin receptor, 2002) VEGFR-3 coreceptor? CD44 Hyaluronan receptor - + (Kriehuber et al., 2001) VEGF-C Growth factor - + (Kriehuber et al., 2001; Hirakawa et al., 2003) VEGFR-1 Growth factor receptor - + (Hirakawa et al., 2003) Neuropilin-1 Semaphorin and growth factor receptor - + (Hong et al., 2002; Petrova et al., 2002) Endoglin/CD105 Low-affinity receptor for TGF-b - ++ (Hirakawa et al., 2003) CD34 L-selectin receptor - ++ (Young et al., 1995) /(+)3 IL-8 CXC-chemokine - + (Petrova et al., 2002) N-cadherin Adhesion molecule - + (Petrova et al., 2002; Hirakawa et al., 2003) ICAM-1/CD54 Adhesion molecule - + (Erhard et al., 1996) Integrin α5 Adhesion molecule, subunit of fibronectin - + (Petrova et al., 2002; Hirakawa et al., receptor 2003) Collagen IV Extracellular matrix protein - ++ (Hirakawa et al., 2003) /(+)4 Versican Chondroitin sulfate proteoglycan - + (Petrova et al., 2002; Hirakawa et al., 2003) Laminin Basement membrane molecule - ++ (Barsky et al., 2003; Petrova et al., /(+)4 2002) Collagen XVIII Basement membrane molecule - ++ (Petrova et al., 2002; Hirakawa et al., /(+)4 2003) PAL-E Caveolae-associated glycoprotein? - ++ (Schlingemann et al., 1985; Niemela et al., 2005) BV, blood vessel; CCL, CC chemokine ligand; LV, lymphatic vessel; LYVE-1, lymphatic vascular endothelial hyaluronan receptor-1;PAL-E, pathologische anatomie leiden-endothelium; VEGF, vascular endothelial growth factor; VEGFR, vascular endothelial growth factor receptor. 1 VEGFR-3 expression was also found on some blood capillaries during tumor neovascularization and in wound granulation tissue (Valtola et al., 1999; Paavonen et al., 2000). 2 Neuropilin-2 is also expressed in veins (Yuan et al., 2002). 3 CD34 expression has also been found on lymphatic endothelial cells (Sauter et al., 1998; Kriehuber et al., 2001). 4 Peripheral lymphatic vessels sometimes have an incomplete basement membrane, large collecting vessels have a complete one.
2.1.2.3 Key lymphangiogenic growth factors
2.1.2.3.1 VEGF-A
25
Vascular endothelial growth factor A (VEGF-A) is an important signaling protein involved in both vasculogenesis - the de novo formation of the embryonic circulatory system - and angiogenesis - the growth of blood vessels from pre-existing vasculature. VEGF-A activity has been mostly studied in cells of the vascular endothelium, although it also exerts effects on a number of other cell types such as monocytes/macrophages and neurons (Liu et al, 2007). In vitro, VEGF-A has been shown to prevent cell apoptosis and to stimulate endothelial cell proliferation, migration, sprouting and tube formation (Ferrara et al, 2003). In vivo, the pro-survival effects of VEGF are developmentally regulated. VEGF inhibition results in extensive apoptotic changes in the vasculature of neonatal but not adult mice (Gerber et al, 1999). VEGF-A is also a vasodilator and increases microvascular permeability and, thus, was originally identified as vascular permeability factor (Senger et al, 1983).
VEGF-A binds to VEGFR-1 (FLT1) and VEGFR-2 (FLK1) as well as to the non- kinase receptors neuropilin-1 (NRP1) and NRP2 (Neufeld et al, 1999; Neufeld et al, 1994). Besides its prominent activity on endothelial cells, VEGF-A also induces hematopoietic stem cell mobilization from the bone marrow, monocyte chemoattraction, osteoblast-mediated bone formation and neuronal protection, since these cell types express VEGF receptors (Ferrara et al, 2003; Storkebaum et al, 2004). Moreover, VEGF-A stimulates the recruitment of inflammatory cells such as macrophages and leads to the expression of proteases implicated in pericellular matrix degradation during angiogenesis (Mandriota et al, 1995; Unemori et al, 1992). VEGF- A expression is strongly induced by hypoxia inducible factor (HIF) under hypoxic conditions as well as by many cytokines including platelet-derived growth factor (PDGF), epidermal growth factor (EGF), basic fibroblast growth factors (FGF-2) and transforming growth factors-α (TGFA) (Cao et al, 2002; Detmar et al, 1994; Dvorak et al, 1995; Ferrara, 2004; Pugh & Ratcliffe, 2003; Wu et al, 2000).
The human VEGF-A gene is organized into eight exons (Houck et al, 1991; Tischer et al, 1991). At least nine VEGF-A isoforms of variable amino acid number are produced through alternative splicing: VEGF121, VEGF145, VEGF148, VEGF162, VEGF165, VEGF165b, VEGF183, VEGF189 and VEGF206 (Bates & Harper, 2002; Lange et al, 2003). VEGF121, VEGF165 and VEGF189 are the major forms secreted
26 by most cell types (Robinson & Stringer, 2001). Among the most commonly observed isoforms, VEGF121 does not bind to heparin and diffuses relatively freely in tissues. In contrast, VEGF189 is sequestered in the extracellular matrix. Enzymatic processing of VEGF189 generates an active form (VEGF110) which lacks the heparin-binding domain (Plouet et al, 1997). This leaves VEGF165 as the most widespread and abundantly expressed splice variant that interacts with heparin sulfate proteoglycans (HSPGs) and neuropilins in a biologically active form. Isoform- specific VEGF-A knockout mice revealed different biological functions of VEGF-A isoforms. Notably, retinal vascular development was normal in mice exclusively expressing the VEGF164 isoform (VEGF164/164), indicating that this isoform contains all necessary information for normal outgrowth and remodeling of blood vessels. In contrast, VEGF120/120 mice exhibited severe vascular defects, with impaired venous and severely defective arterial vascular development in the retina. VEGF188/188 mice had normal venous development, but aborted arterial outgrowth (Carmeliet et al,
1999; Stalmans et al, 2002). Transgenic mice overexpressing VEGF164 under the K14 promoter showed a psoriatic phenotype with distinctive vascular changes, epidermal alterations, and inflammatory infiltrates closely resembling human psoriasis (Xia et al, 2003). Mice with a targeted deletion of VEGF in the skin exhibit delayed wound healing and less frequent development of chemically induced papillomas (Rossiter et al, 2004).
2.1.2.3.2 VEGFR-2
VEGFR-2, also known as KDR or FLK1, is a receptor for VEGF-A, VEGF-C and VEGF-D (only in human), and is expressed by both blood vascular and lymphatic endothelium (Hirakawa et al, 2003; Hong et al, 2004b; Kriehuber et al, 2001). The role of VEGFR-2 in angiogenesis has been thoroughly examined, however, the function of VEGF-A signaling through VEGFR-2 in lymphangiogenesis still needs to be further elucidated. Cell proliferation assays demonstrated that VEGF-A potently induced proliferation of lymphatic endothelial cells in vitro (Hirakawa et al, 2005b). Additionally, injection of adenoviral murine VEGF-A164 demonstrated pronounced and recurrent in vivo lymphangiogenesis in mouse ears (Nagy et al, 2002). Conversely, adenovirus expressing the human VEGF-A165 isoform did not show
27 distinct lymphangiogenic activity in mouse models (Byzova et al, 2002; Enholm et al, 2001). A likely explanation for this different phenotype may be due to species- specific effects of VEGF-A or due to the tissue-specific lymphangiogenic potency of VEGF-A. Furthermore, Hong et al. and Kunstfeld et al. have recently revealed that skin-specific overexpression of murine VEGF-A164 resulted in enhanced lymphangiogenesis during tissue repair and in skin inflammation, respectively (Hong et al, 2004b; Kunstfeld et al, 2004). In addition, Hong et al. further demonstrated that blocking VEGFR-2 signaling by a VEGFR-2 blocking antibody inhibited both angiogenesis and lymphangiogenesis in healing wounds, indicating the importance of VEGFR-2 for repair-associated lymphangiogenesis (Hong et al, 2004b).
2.1.2.3.3 Angiopoietin-2
Angiopoietin-2 (Ang2) is a ligand for the endothelial cell-specific tyrosine kinase receptor Tie2 and likely acts as an antagonist for angiopoietin-1 (Ang1) (Maisonpierre et al, 1997; Suri et al, 1996). Ang2 destabilizes interactions between blood vascular endothelial cells and surrounding pericytes, resulting in diminished endothelial cell - pericyte contacts. In contrast, Ang1, which interferes with the Ang2 signaling, stabilizes mature blood vessels (Thurston, 2003). Ang2 may induce angiogenic sprouting in the presence of VEGF but may stimulate vessel regression in the absence of VEGF (Holash et al, 1999a; Holash et al, 1999b; Thurston, 2003; Whitehurst et al, 2007). Ang2-deficient mice displayed chylous ascites, lymphedema, and lymphatic dysfunction but replacement of Ang2 with Ang1 was sufficient to rescue the lymphatic vascular phenotype. Although Ang2 is not required during early lymphatic vessel formation, large lymphatic vessels of Ang2 mutant mice were structurally irregular and leaky, and smaller lymphatic vessels displayed abnormal patterning (Gale et al, 2002).
2.1.2.3.4 HGF
Hepatocyte growth factor (HGF, also known as scatter factor) was recently identified as a potent lymphangiogenesis factor (Kajiya et al, 2005). HGF binds directly to its
28 receptor HGF-R to induce proliferation, migration and tube formation of LEC in vitro, exerting its effects independently of the VEGFR-3 pathway. HGF additionally interacts with integrin alpha 9, expressed specifically on LEC, to promote migration. HGF transgenic mice have increased numbers and enlargement of lymphatic vessels and similar results are observed when HGF is delivered subcutaneously, demonstrating that HGF can directly promote lymphangiogenesis in vivo.
2.1.2.3.5 FGF2
The role of fibroblast growth factor-2 (FGF2) in vascular development and angiogenesis has been well characterized (Auguste et al, 2003a). But recently, studies have shown that FGF2 promotes both lymphatic and blood vessel growth in the mouse cornea assay (Chang et al, 2004; Kubo et al, 2002). FGF2 also promotes proliferation and migration of LEC by directly binding to the receptor FGFR-3, which is upregulated by Prox1 in lymphatic endothelium (Shin et al, 2006). The pro- migratory effect could not be abrogated by neutralization of VEGFR-3, suggesting that FGF2 functions independently of the VEGF-C/VEGFR-3 pathway (Shin et al, 2006).
2.1.2.3.6 PDGF/IGF
Recent studies have reported that platelet-derived growth factor-BB (PDGF-BB) and insulin-like growth factor-1 (IGF1) and -2 (IGF2) also induce lymphangiogenesis (Bjorndahl et al, 2005; Cao et al, 2004) in the mouse corneal assay. However, their potential effects on skin lymphangiogenesis still need to be elucidated. Lymphatic vessel formation and growth during physiological and pathological conditions may require interplay of several lymphangiogenic growth factors. Thus, dissecting the molecular mechanisms of these growth factors will provide better insight into understanding lymphangiogenesis in pathologies of the lymphatic vasculature.
29 DGF-8B
HGF-l (0FD ~ \ \ I PDGFRalß VEGFR-2 Nrp2 Tie2 = = =
Lymphatic vessel
Lymphangiogenesis
Figure 2.1.3 Schematic representation of lymphangiogenic growth factors and their receptors expressed by lymphatic endothelium. Several vascular endothelial growth factors (VEGF-A, VEGF- C, VEGF-D) promote lymphangiogenesis by activation of distinct VEGFRs and Nrp2. FGF-2 acts directly through FGFR-3. Angiopoietin-1 (Ang1) activates Tie2 and up-regulates VEGFR-3. HGF, IGF, PDGF-BB, adrenomedullin act directly through their respective receptors HGF-R, IGF-1R, PDGFR and Calcrl/RAMP2.
2.1.3 Pathologies of the lymphatic vasculature
2.1.3.1 Lymphatic Dysfunction, Lymphedema
Abnormal vessel development or damaged lymphatic vessels cause stagnation of proteins and fluid in the interstitium, and lead to lymphedema. This impairment of the lymphatic transport capacity leads to chronic and disabling swelling, tissue fibrosis, adipose degeneration, poor immune function, susceptibility to infections, and impaired wound healing (Rockson, 2001). Primary lymphedemas are rare genetic developmental disorders and are characterized by enlarged lymphatic capillaries and interstitial accumulation of lymph fluid. The symptoms are apparent from birth (Milroy; OMIM:153100) or at puberty (Meige; OMIM:153200) (Witte et al, 1998). Secondary lymphedema is caused by filariasis (elephantiasis) or by trauma due to radiation therapy, surgery or infection. Filariasis is the main cause of lymphedema in tropical countries, with some 100 million people affected worldwide, whereas breast- cancer surgery is a leading cause for secondary lymphedema in industrialized countries (Rockson, 2001). Recent studies have identified mutations in genes that are associated with different human lymphedema syndromes. In Milroy disease, several
30 heterozygous VEGFR-3 missense mutations result in the expression of an inactive tyrosine kinase (Irrthum et al, 2000; Karkkainen et al, 2000). In lymphedema- distichiasis, an autosomal-dominant disorder with congential lymphedema and double rows of eyelashes (distichiasis), inactivating mutations of the FOXC2 gene were identified in several families (Fang et al, 2000). Additionally, Foxc2-targeted mice have lymphatic abnormalities (Fang et al, 2000; Kriederman et al, 2003; Petrova et al, 2004). Patients with FOXC2 mutations display abnormal mural cell coating of their lymphatic vessels and lack lymphatic valves (Petrova et al, 2004). Moreover, mutations of the SOX18 gene on chromosome 20q13, a SRY-related transcription factor, cause recessive and dominant forms of hypotrichosis-lymphedema- telangiectasia syndrome. Mutations in the DNA-binding domain of SOX18 have been found in the recessive form of the disease whereas the dominant hereditary form is caused by a heterozygous nonsense mutation of the transactivation domain (Irrthum et al, 2003). Recently, a possible lymphedema treatment using viral gene-transfer vectors that encode VEGF-C has been reported. VEGF-C gene therapy was effective in Chy mice that suffer from lymphedema caused by a heterozygous inactivating mutation of VEGFR-3. Moreover, VEGF-C156S, which selectively activates VEGFR-3, successfully induced the formation of a functional cutaneous lymphatic vessel network without blood vessel growth or vascular leakiness - side effects observed with VEGF-C gene therapy due to its activation of VEGFR-2 (Saaristo et al, 2002). More recently, successful regeneration of a lymphatic network was observed after transduction of VEGF-C in lymph node transplantation models (Tammela et al, 2007).
2.1.3.2 Lymphatic vessels in inflammation and the immune response
There is increasing evidence that inflammation triggers lymphangiogenesis mediators that may regulate lymphatic vessel function. Lymphatic vessels participate in the regulation of inflammatory responses through their role in the transport of lymphocytes to lymph nodes. Migration of dendritic cells is mediated by the chemokine receptor CCR7 whereas lymphatic vessels express the ligand CCL21 (Ohl et al, 2004). Furthermore, inflammatory infiltrates in human kidney transplants undergoing rejection contain proliferating host lymphatics (Kerjaschki et al, 2004),
31 and lymphangiogenesis has also been observed in experimental models of chronic airway inflammation (Baluk et al, 2005). Kajiya and Detmar recently found that acute UVB irradiation of the skin results in hyperpermeable, leaky lymphatic vessels that are functionally impaired, and that blockade of VEGFR-3 resulted in prolonged inflammation and edema after UVB irradiation (Kajiya & Detmar, 2006).
Psoriatic skin lesions are characterized by pronounced lymphatic hyperplasia and an increase in the numbers and size of blood vessel (Kunstfeld et al, 2004). The pathogenesis of psoriasis remains unclear, although it is generally accepted that activated T lymphocytes and dendritic cells are important in the maintenance of psoriasis (Gottlieb et al, 2005). Although psoriasis appears to be a human-specific disease, homozygous VEGF transgenic mice spontaneously develop psoriasis-like inflammatory skin lesions at around 6 months of age. Heterozygous VEGF transgenic mice do not develop inflammatory skin lesions, but only upon induction of delayed type hypersensitive reaction (DTH) with oxazolone (Kunstfeld et al, 2004).
2.1.3.3 Lymphatic involvement in tumor metastasis
Metastatic tumor spread to regional lymph nodes through lymphatic vessels is the most important prognostic factor for tumors of epithelial origin (Dadras et al, 2005). At present, little is known about the mechanisms how tumor cells gain entry into the lymphatic system and increase their potential for subsequent organ metastasis. The sentinel lymph node is the first regional lymph node to which tumor cells metastasize and it is important in the staging, treatment, and follow up of many solid tumors (Pepper, 2001).
The process of metastasis through the lymphatic vessel system is complex and involves changes in the expression of numerous genes (Ramaswamy et al, 2003). Growth factor-mediated stimulation of lymphatic vessels appears to be required for lymphatic metastasis. Recent studies in animal tumor models have provided evidence that increased levels of VEGF-C and/or VEGF-D promote active tumor lymphangiogenesis and lymphatic tumor spread to regional lymph nodes, and that these effects can be suppressed by blocking VEGFR-3 signaling (Mandriota et al,
32 2001; Skobe et al, 2001; Stacker et al, 2001). Tumor cells and tumor-associated macrophages secrete VEGF-C and VEGF-D, which induce sprouting of nearby lymphatic vessels, facilitating the egress of tumor cells into the vessel lumen. The lymphatic endothelial cells may also actively attract some tumor cells through the secretion of chemokines, such as CCL21 (Kriehuber et al, 2001). VEGF-A has been shown to induce VEGF-C expression in cultured BEC, and VEGF-A producing transgenic tumors indeed showed higher VEGF-C protein levels than wild-type tumors.
Traditionally, VEGF-A was known as a blood vessel-specific growth factor. However, its major signaling receptor, VEGFR-2, is also expressed on lymphatic endothelial cells in vitro and in situ (Hirakawa et al, 2003; Hong et al, 2004b; Kriehuber et al, 2001; Saaristo et al, 2002). A high correlation between VEGF-A production and lymph node metastasis in several cancer types, including gastric cancer have been described (Hirakawa et al, 2005b; Kimura et al, 2001). Mice overexpressing VEGF-A in the skin, subjected to a standard chemically-induced skin carcinogenesis regimen, showed active proliferation of VEGFR-2-expressing tumor- associated lymphatic vessels, as well as enhanced tumor metastasis to the sentinel and distant lymph nodes (Fig. 2.1.4) (Hirakawa et al, 2005b; Tobler & Detmar, 2006). The pioneering finding of this study was that even before metastasizing, VEGF-A overexpressing primary tumors induced sentinel lymph node lymphangiogenesis, preparing their future metastatic spread (Hirakawa et al, 2005b). The relative contribution of direct (via activation of VEGFR-2 on LECs) versus indirect effects toward the lymphangiogenic activity of VEGF-A remains to be explored.
33 IX_= ......
Figure 2.1.4 VEGF-A expressing cancer cells induce tumor and lymph node lymphangiogenesis. In normal skin, lymphatic vessels are present in the dermis and maintain tissue fluid homeostasis. There is no detectable lymphangiogenesis within draining LN. SCC of K14/VEGF-A transgenic mice induce primary, tumor-associated, lymphatic vessel growth but also lymphangiogenesis within sentinel LN, even before they metastasize, possibly preparing the LN for their later arrival. Metastatic, VEGF- A expressing SCC maintains their lymphangiogenic activity after metastasis to sentinel LN (Tobler & Detmar, 2006).
Recent reports have demonstrated that leukocytes also play an important role in promoting tumor-associated lymphatic vessel growth and activation. Activated macrophages express VEGFR-3, and the lymphangiogenic factor VEGF-C has been shown to enhance macrophage chemotaxis (Schoppmann et al, 2002; Skobe et al, 2001) whereas tumor-associated macrophages secrete VEGF-C to promote lymphangiogenesis (Schoppmann et al, 2002). Recently, VEGF-A secreted by follicular B cells has been implicated in the mediation of lymph node lymphangiogenesis (Angeli et al, 2006), however, the relative contribution of leukocyte-derived lymphangiogenic factors needs to be further elucidated. Bone
34 marrow-derived progenitor cells and macrophages may physically incorporate into newly formed lymphatic vessels (Kerjaschki, 2005; Maruyama et al, 2005; Schledzewski et al, 2006), but this has not been observed in tumor-associated lymphangiogenesis (He et al, 2004), and it remains unclear whether this mechanism significantly contributes to pathological lymphangiogenesis.
Therapeutically, a soluble VEGFR-3/Fc molecule ("VEGF-C/D trap") inhibited the sprouting and lymphatic vessel enlargement and seemed to restore the integrity of the lymphatic vessel wall (He et al, 2005). Similarly, blocking monoclonal antibodies against VEGF-C/-D or their receptor(s), and small molecules that inhibit the tyrosine kinase catalytic domain of these receptors could be used for the inhibition of lymphangiogenesis and hence tumor metastasis.
2.2 MICROARRAY TECHNOLOGY
The introduction of microarray technology has allowed researchers to examine various biological questions on a genome-wide scale. It has provided a systematic way to study gene expression across the entire transcriptome. The transcriptome is the complete set of transcripts and their relative levels of expression in a particular cell or tissue type under defined conditions (Schena et al, 1995; Shalon et al, 1996).
2.2.1 Oligonucleotide microarray technology
Unlike cDNA microarrays that use expression sequence tags (ESTs) extracted from a sequenced cDNA library, oligonucleotide microarrays contain a series of 25-mer (in case of the Affymetrix platform) or 60-mer (in case of the Applied Biosystems platform) oligonucleotides designed by a computer algorithm to represent known or predicted open reading frames (Kuo et al, 2006). Target labeling is performed using amplified RNA (aRNA) rather than cDNA. The first-strand reverse transcription of poly-A mRNA is performed as for cDNA microarrays, but the poly-dT primer includes a promoter sequence for the enzyme T7 RNA polymerase. After synthesis of the second strand, the T7 enzyme is added and it synthesizes multiple copies of
35 antisense RNA of the gene, incorporating biotinylated (Affymetrix, http://www.affymetrix.com) or DIG-oxigenin (Applied Biosystems, http://www.appliedbiosystems.com) nucleotides during the reaction (Bammler et al, 2005).
Hybridization to oligonucleotide arrays is a noncompetitive, single-colored per array- approach, and is detected by addition of either a fluorescently labeled streptavidin compound that binds to the biotin group in the aRNA molecule (Affymetrix) or a chemiluminescently labeled anti-digoxigenin-alkaline phosphotase compound that can be detected with digoxigenin in the aRNA molecule (Applied Biosystems).
Some of the advantages of oligonucleotide arrays include the accommodation of higher densities of genes and a lower variability from chip to chip. Researchers can also use this approach without access to microarray construction facilities, and these arrays lend themselves to data comparison across research groups. However, cDNA microarrays are considerably cheaper and offer higher levels of replication that in turn promotes statistical analysis (Wei et al, 2004). They also rely on hybridization over kilobases rather than tens of bases, which may reduce cross-hybridization artifacts, and also minimize effects of intra-specific single nucleotide polymorphisms on the hybridization that could be misinterpreted as strain-specific variation of gene expression (Patterson et al, 2006).
2.2.2 Gene expression data analysis
Microarray technology measures the relative amount of mRNA expressed in two or more experimental conditions (e.g. no-treatment vs. treatment, healthy vs. disease) and generates a differential expression of all known genes (approximately 30,000 genes in humans). Despite the comprehensiveness of microarray technology, one of the main challenges of gene expression profiling is data analysis. The semi- quantitative measurements from microarrays lend themselves to the generation of false positive and false negative results (Breitling, 2006; Couzin, 2006; Magic et al, 2007). In order to circumvent these errors, one can select differentially expressed genes by applying a variety of statistical tests to consider both fold-change and
36 variability in combination to create a p-value, an estimate of how frequently we would observe these data by chance alone (Chen, 2007; Vardhanabhuti et al, 2006). While the statistical analyses may reliably identify which gene products are differentially expressed under different experimental conditions, making a significant biological extrapolation of differentially expressed genes is key to understand biological effects using microarrays (Breitling, 2006). Having identified some set of differentially expressed genes, the subsequent analysis involves identification of biological patterns within the given set of genes. Gene ontology analysis provides a standard way to define these relationships (Dai et al, 2005; Subramanian et al, 2005). Most genes but not all have several key attributes involved in biological processes, molecular functions and biological pathways. Categorization of regulated genes based on their gene ontology terms generates important relationships between genes and a given condition and thus may generate new hypotheses for further biological investigations (Curtis et al, 2005; Subramanian et al, 2005).
Expression profiling using microarray technology provides exciting new information about what genes do under various conditions. However, the size and complexity of these experiments often results in a wide variety of possible interpretations. Therefore, good experimental design, adequate biological replication and follow-up experiments play key roles in the successful extrapolation of profiling experiments (Couzin, 2006).
37 3 RESULTS AND DISCUSSION
38 3.1 Prox1 promotes lineage-specific expression of FGF receptor-3 in lymphatic endothelium
3.1.1 Introduction
Fibroblast growth factor (FGF) signaling plays an important role in a broad range of biological processes of vascular endothelial cells including proliferation, migration, survival, tubulogenesis and differentiation (Javerzat et al, 2002). At least, twenty three different FGFs and four FGF-receptors (FGFR-1 through FGFR-4) have been identified and characterized in vertebrates so far (Javerzat et al, 2002; Ornitz & Itoh, 2001). FGFRs belong to the receptor tyrosine kinase family and commonly consist of three extracellular immunoglobulin-like domains, a single-pass transmembrane domain and a split-tyrosine kinase domain. Alternative splicing generates a wide array of isoforms of FGFRs with distinct physical and biological characteristics (Dell & Williams, 1992; Groth & Lardelli, 2002; Hanneken, 2001; Ornitz, 2000; Terada et al, 2001; Wilkie et al, 2002). The most common variants, the IIIb or IIIc isoform, are formed by alternative splicing of the carboxy-terminal half of the third immunoglobulin domain of FGFR-1, -2 and -3, but not FGFR-4. The alternative splicing is regulated in a tissue-specific manner and also determines their binding specificity for various FGF ligands. In general, the IIIb isoforms of FGFRs are predominantly expressed by epithelial lineage cells, whereas the IIIc variants tend to be expressed in mesenchymal lineages (Alarid et al, 1994; Murgue et al, 1994; Orr- Urtreger et al, 1993; Yan et al, 1993). FGF ligands and their interacting receptor isoforms are often expressed in adjacent tissues.
The roles of FGFs in vascular development have been well characterized in the context of angiogenesis that is associated with tumor development, tissue repair and embryogenesis (Auguste et al, 2003b; Bikfalvi et al, 1998; Javerzat et al, 2002). FGF- 2 was one of the first angiogenic factors identified for its potent activity on vascular endothelial cell proliferation (Shing et al, 1984). Recently, FGF-2 was reported to also induce lymphatic vessel growth in mouse cornea assay by promoting the secretion of the potent lymphangiogenic factor, vascular endothelial cell growth factor (VEGF)-C, by blood vascular endothelial cells (Chang et al, 2004; Kubo et al, 2002). Moreover,
39 systemic treatment with a blocking antibody against VEGFR-3, the major receptor for VEGF-C, reduced the FGF-2-induced corneal lymphangiogenesis (Chang et al, 2004; Kubo et al, 2002). These findings indicate that the effects of FGF-2 on lymphangiogenesis might be largely indirect through activation of the VEGF- C/VEGFR-3 signaling pathway.
The homeodomain transcriptional factor Prox1 was originally isolated due to its homology with the Drosophila Prospero protein (Oliver et al, 1993b). Like Prospero, Prox1 plays an important role in cell fate decisions of diverse cell types and serves as a master regulator during embryonic development of the lymphatic vascular system (Hong et al, 2002; Wigle et al, 2002b; Wigle & Oliver, 1999). Upon an inductive signal during early development, Prox1 is upregulated in a subset of venous endothelial cells and reprograms their gene expression profile similar to that of lymphatic endothelial cells. The Prox1-positive venous endothelial cells then further differentiate to adopt lymphatic endothelial cell phenotypes and migrate out to form the primitive lymphatic vessels. Therefore, the Prox1-mediated cell fate reprogramming is the initial and essential step during lymphatic endothelial differentiation (Wigle et al, 2002b; Wigle & Oliver, 1999). In addition, we and others have recently found that ectopic over-expression of Prox1 in cultured blood vascular endothelial cells (BECs) isolated from human foreskin recapitulates the embryonic lymphatic reprogramming by down-regulating the BEC-specific genes and by up- regulating several lymphatic-specific genes (Hirakawa et al, 2003; Hong et al, 2002; Petrova et al, 2002b). However, the molecular mechanisms underlying this lymphatic reprogramming are poorly understood. In this study, we present evidence demonstrating that Prox1 upregulates the expression of FGFR-3 during lymphatic reprogramming and that FGF signaling through the upregulated FGFR-3 plays an important role in the early lymphatic vascular system development.
3.1.2 Results
3.1.2.1 Ectopic expression of Prox1 in primary BEC upregulates FGFR-3
40 We and others have previously reported that ectopic expression of Prox1 in blood vascular endothelial cells (BECs) led to upregulation of several LEC-specific genes (Hong et al, 2002; Petrova et al, 2002b). Detailed microarray analyses further indicate that expression of FGFR-3 is also regulated by the expression of Prox1 in BECs. Real-time RT-PCR analyses confirmed that Prox1 increased FGFR-3 expression by 20-fold (Fig. 3.1.1A). The Prox1-mediated upregulation of FGFR-3 was further confirmed by Northern blot analysis by using total RNAs harvested 3, 24 and 48 hours after transduction of BECs with an adenovirus expressing Prox1 (Fig. 3.1.1B).
To determine which of the two major FGFR-3 isoforms (IIIb and IIIc) was upregulated by Prox1, RT-PCR was performed by using primers designed to yield a 235-bp product containing an ApaI site from the IIIb isoform, or a 229-bp fragment without an ApaI site from the IIIc isoform. As controls for the analyses, we also used stably transfected myoblast cells that selectively express either human FGFR-3 IIIb or IIIc isoform (Kanai et al, 1997). As expected, RT-PCR analysis yielded an ApaI- sensitive 235-bp product from the IIIb-expressing control cells, and an ApaI- insensitive 229-bp fragment from the IIIc-expressing cells (Fig. 3.1.1C). The same analysis amplified an ApaI-insensitive 229-bp product from BECs infected with the Prox1-adenovirus (Fig. 3.1.1C). This indicates that Prox1 predominantly upregulates expression of the FGFR-3 IIIc isoform with a slight activation of the expression of IIIb is also in vascular endothelial cells. Furthermore, RT-PCR of RNA obtained from primary lymphatic endothelial cells generated the ApaI-insensitive 229-bp product, whereas unpurified cell mixtures isolated from human foreskins yielded products of both ApaI-sensitive and insensitive fragments (Fig. 3.1.1D). These data indicate that the FGFR-3 IIIc isoform is the major variant present in LECs and that Prox1 selectively upregulates the IIIc isoform of FGFR-3.
41 3h 24 h 48 h A 3000 B C p CP CP .FGFR-3 2500 Proxl ;:; 2000 ;> - 18S j
~ 1500 1:: FGFR3 ~ t.l 1000
500 28S rR A 0 AdCTR AdProxl C R3/Proxl Illb IIIc D U C U C U c S LI L2 IIIc u CC U C U C - 229 bp - 229 bp - 169 bp - 169 bp 1 2 3 4 5 6 7 8 2 3 4 5 6
E 400 350
Gi 300 > .E! 250 0 ·iii" 11I 200 ~ Co )( 150 w 100
50
0 AdCTR AdProxl AdmutProxl
Figure 3.1.1 Prox1 upregulates FGFR-3 expression. (A) The steady-state level of FGFR-3 mRNA was increased by 20-fold when Prox1 is ectopically overexpressed in BECs. FGFR-3 expression level was measured by real-time RT-PCR in BECs after transduced with control (AdCTR) or Prox1 (AdProx1) adenovirus. Data were normalized by β-actin mRNA levels and expressed as % of the control virus-infected cells (means±SD). (B) Upregulation of FGFR-3 mRNA expression by Prox1 was confirmed by Northern blot analysis of total RNAs obtained from BECs infected with control (C) or Prox1 (P) adenovirus for 3, 24 or 48 hours. (C) Prox1 induces expression of the IIIc isoform of FGFR- 3, as determined by a diagnostic ApaI restriction analysis of RT-PCR product (229-bp) amplified from BECs transduced with Prox1-adenovirus for 24 hours (R3/Prox1). As controls, RT-PCR products from FGFR3 IIIb or IIIc-expressing cell lines were digested in parallel. Only the product from the IIIb isoform was digested to yield a 169-bp fragment. (D) Cultured lymphatic endothelial cells (LECs) exclusively express the IIIc isoform of FGFR-3. RT-PCR products of unpurified cell mixture (S) from human neonatal foreskins, two independent batches of LECs (L1, L2) and FGFR-3 IIIc-expressing control cells (IIIc) were subjected to the diagnostic ApaI restriction analysis. While the product from the cell mixture contains both ApaI sensitive and resistant fragments, those of LECs and of FGFR-3 IIIc control cells (IIIc) were resistant to the digestion, indicating that the IIIc is the dominant isoform of FGFR-3 in LECs. U: undigested, C: digested with ApaI. (E) Cultured lymphatic endothelial cells were transduced with adenovirus expressing the wildtype (AdProx1) or mutant (AdmutProx1) Prox1, or with control adenovirus (AdCTR). After 2 days, the expression level of FGFR-3 was determined using real-time RT-PCR.
42 To determine if Prox1 is necessary to maintain the expression of FGFR-3 in LECs, we ectopically expressed a mutant Prox1 in cultured lymphatic endothelial cells through the adenovirus gene transfer. The mutant Prox1 protein has two amino-acid substitution mutations in its DNA-binding domain and does not display any transcriptional activity (see below). We found that when expressed in LECs, the mutant Prox1 was able to decrease the expression level of FGFR-3 by 4-fold, while the wildtype Prox1 upregulated FGFR-3 by 3-fold (Fig. 3.1.1E). These findings indicate that the mutant Prox1, serving as a dominant negative mutant, may compete with the endogenous Prox1 in LECs and that Prox1 function is necessary to maintain the expression of FGFR-3.
3.1.2.2 Prox1 binds to the FGFR-3 promoter and activates its transcription
To study the molecular mechanisms underlying the Prox1-mediated upregulation of FGFR-3, we performed promoter-reporter assays using FGFR-3 promoter-luciferase constructs, which have been previously characterized (McEwen & Ornitz, 1998). A 3- kb promoter fragment was sufficient to mediate transcriptional activation of the firefly luciferase reporter (P1) by Prox1 (Fig. 3.1.2A). The Prox1-mediated activation was still maintained even after deleting most of the promoter region to –220 n.t. upstream of the FGFR-3 transcriptional initiation site (P2), but removal of the proximal 220-bp of the promoter sequence abrogated the activation by Prox1 (P3). The Prox1-mediated activation progressively decreased with deletions to –175 and –126 n.t. and then was completely abolished by deletion to -79 n.t. (P4-P6). These data indicate the presence of putative Prox1 response elements (PRE) between -220 to -79 n.t. of the FGFR-3 promoter.
43 A luciferase Atitvity -2951 -27 " " '------1~ PI
·1537 ·220 ''------1~ DCtrl
[lProlCl
.MutProx1
~pGl.Z
~-" ::S~ ~;;::!2======:lI-~ ~ ::S~ 1:::==:>-< B CrTNNCr CACHNCr -190 111I111 1111111 accCrTCCCT<:gce tCjc tccqceccAgC t 999CtecCACGCCIC'1"999 111 11I1 CGTCTTn caTeTTa
CACNNCC CACNNCT I 111 I 1111111 -100 accgcec99C'9ccccCGCCTOACCACGCCTCTTC9911tCtCC 1111 I 11111 CGTen. caTe'I'Ta c mPGf'R3:
_160 _100 TCi1i1I1I1C t li g9ctccCACGCCCTCGAgaccgccgggcgCCCCCOCCCGGcCACGCCCCCTcggatgcccnnllliliIl 1111II1I1i1ii"llilililll11irc -208 -148
hP'GP'R):
1 2 J , 5 6 7 D Probt' WT WT WT WT i\1T ~1T MT GST GST-ProxßD + + + COlllpClilor +
Frcc [ Probe
Figure 3.1.2 Prox1 binds to the FGFR-3 promoter to upregulate its transcription. (A) The FGFR- 3 promoter-luciferase (Luc) reporter constructs (P1-P6) and an empty control vector (pGL2) were tested for their luciferase activity in the presence of a vector control (Ctrl), a Prox1-expressing vector (Prox1), or a mutant Prox1-expressing vector (MutProx1). P6-6XPBS contains six-tandem repeats of the Prox1 binding site (PBS, CACGCCTCT) in the P6 construct in the forward- (P6-6XPBS_F) or reverse (P6-6XPBS_R) orientation. Numbers indicate relative locations from the transcriptional initiation site (McEwen & Ornitz, 1998). Data are shown as means±SD. (B) Sequence analysis of the mouse FGFR-3 promoter region revealed four putative Prox1 binding sites. Two previously reported Prospero consensus sequences, C(A/T)(C/T)NNC(T/C) and CGTCTT(A) (Cook et al, 2003; Hassan et al, 1997), are shown above and below the putative Prox1 binding sites (bold), respectively. (C) The proximal three putative Prox1 binding sequences (bold) are conserved between the mouse and human FGFR-3 promoters. (D) Gel electrophoresis mobility shift assays showed that the purified GST-Prox1 fusion protein (GST-ProxBD), but not the GST protein alone, binds to the putative Prox1 binding sequences found in the FGFR-3 promoter. The GST or GST-Prox1 proteins were incubated with 32P- labeled a wild type (WT) or a mutant probe (MT). Arrow indicates a slow migrating complex of GST- Prox1 and the wild-type probe (lane 3). Excessive amount of unlabeled wild type probe (Competitor) competes for the interaction between GST-Prox1 and the labeled wild-type probe (lane 4).
44
To further determine whether the Prox1-mediated activation of FGFR-3 is dependent on DNA-protein interaction, we introduced two amino acid substitution mutations (N625A and R627A) into the third helix of the homeodomain region that is involved in DNA binding of Prospero, the Drosophila homolog of Prox1 (Ryter et al, 2002). Prospero and Prox1 share a high amino acid identity in their DNA-binding domains (Hong & Detmar, 2003b). The Prox1 protein with the two substitution mutations (MutProx1) completely lost its transcriptional activity (Fig. 3.1.2A). These findings indicate that a direct DNA-protein interaction is necessary for the Prox1-mediated upregulation of FGFR-3.
3.1.2.3 Identification of the putative Prox1 binding sites in the FGFR-3 promoter
Previous reports had identified two seemingly different consensus sequences (C(a/t)(c/t)NNC(t/c) and (T)AAGACG) as putative Prospero binding sites (Cook et al, 2003; Hassan et al, 1997). Interestingly, we found four putative Prox1 binding sites, composed of the two partially overlapping Prospero consensus sites, between -190 and -100 n.t. of the mouse FGFR-3 promoter (Fig. 3.1.2B). The proximal three putative Prox1 binding sequences are highly conserved between the mouse and human FGFR-3 promoters (Fig. 3.1.2C). To investigate whether these sequence motifs serve as Prox1 binding sites, we performed gel electrophoresis mobility shift assays (EMSA) using a GST-Prox1 fusion protein (Belecky-Adams et al, 1997; Cui et al, 2004). Purified GST-Prox1 fusion protein efficiently bound to a probe containing the putative Prox1 site in the FGFR-3 promoter (Fig. 3.1.2D). However, the fusion protein did not bind to a mutant probe whose putative Prox1 site was replaced with random nucleotides. Interaction of the GST-Prox1 protein with the labeled wild-type probe was competed out by addition of excessive unlabeled wild type probe and the GST protein alone did not interact with either probe (Fig. 3.1.2D). These data demonstrate that Prox1 bind to the putative Prox1 site present in the FGFR-3 promoter.
We next investigated whether the Prox1 binding site identified in the FGFR-3 promoter is sufficient to mediate transcriptional activation of the reporter gene. We introduced six tandem repeats of the Prox1 binding sequences (PBS, CACGCCTCT)
45 into the P6 construct in the forward- or the reverse orientation (P6-6XPBS_F and P6- 6XPBS_R) (Fig. 3.1.2A). The P6 construct was shown to be unable to mediate any transcriptional activation by Prox1. However, introduction of six repeats of the putative Prox1 binding sites in the forward orientation (P6-6XPBS_F) re-instated transcriptional activity to the P6 construct by wild-type, but not by the mutant Prox1 (Fig. 3.1.2A). However, when the repeats were introduced in the reverse orientation (P6-6XPBS_R), only marginal activation was observed. These findings indicate that the nine-nucleotide sequence (CACGCCTCT) present in the FGFR-3 promoter is necessary and sufficient to mediate transcriptional activation by Prox1.
3.1.2.4 Expression of FGFR-3 in developing lymphatic vessels of mouse embryo and of human skin
We next investigated whether FGFR-3 is expressed in the lymphatically differentiating endothelial cells during mouse embryogenesis. In agreement with our in vitro results, many of the Prox1-positive lymphatically differentiating endothelial cells were positively stained for FGFR-3 in E11.5 mouse embryos (Fig. 3.1.3A-D). Furthermore, double immunofluorescent stainings for the lymphatic-specific marker LYVE-1 and for FGFR-3 showed that FGFR-3 was strongly and specifically expressed in the newly formed LYVE-1-positive lymphatic vessels (Fig. 3.1.3E-H), but not in developing blood vessels (cardinal vein) (Fig. 3.1.3C and G). Furthermore, double stainings of human neonatal foreskin for LYVE-1 and FGFR-3 revealed that the lymphatic specific expression of FGFR-3 is also maintained after embryonic development (Fig. 3.1.3I-L).
46
Figure 3.1.3 FGFR-3 expression in lymphatic endothelial cells during and after embryonic development. Adjacent mouse embryo sections (E11.5) were stained for Prox1 (green) and FGFR-3 (red) (A-D), and for LYVE-1 (green) and FGFR-3 (red) (E-H). Lymphatically differentiating budding endothelial cells and resident endothelial cells in a newly formed lymphatic vessel are co-stained positively for Prox1 and for FGFR-3 (D). Similarly, LYVE-1-positive lymphatic endothelial cells express FGFR-3 (H). Arrows indicate a newly formed lymphatic vessel (B-D and F-H). A human neonatal foreskin section was co-stained for LYVE-1 and FGFR-3 (I-L). Asterisk, cardinal vein; bar, 100 µm.
3.1.2.5 Signaling through FGFR-3 promotes LEC proliferation
To further evaluate the biological role of FGFR-3-mediated signaling, we inhibited expression of FGFR-3 in LECs using small interfering RNAs (siRNAs) and studied the effects on cell proliferation. Real-time RT-PCR analyses revealed that transfection of FGFR-3 siRNAs into LECs decreased the steady state level of FGFR-3 by 50-fold, whereas the expression of FGFR-1 was not altered (Fig. 3.1.4A). Notably, knockdown of FGFR-3 resulted in a significant inhibition of proliferation of LECs by 30-40% (Fig. 3.1.4B). However, the FGF-2-induced proliferation of LECs was largely unaffected (about two-fold) with or without inhibition of FGFR-3. This may be due to the presence of other functional FGF receptors (FGFR-1, -2 and -4) that
47 may be activated by FGF2. Together, these data demonstrate that the FGFR-3- mediated signaling plays an important role in proliferation of LECs.
A 140 n.5. """ 120 11 11
100
=0 '<;j 80 CI> ...01 Q" ~ 60 01 Oll S =40 ...01 ""01 c.. 20
0 FGFRI FGFR3 """ B " 3500 """ "" oNone 3000 oFGF2 ,,"Heparin 2500 • FGF2/heparin tl '2 """ """ ;:J 2000 11 11
01 ==CI> ""01 ... 1500 0 ii:= 1000
500
siCTR siFGFR3 Figure 3.1.4 FGFR-3 mediates proliferation signaling of LECs. (A) siRNAs against FGFR-3 significantly reduced the steady state levels of FGFR-3, but not of FGFR-1. (B) Knockdown of FGFR-3 inhibits proliferation of LECs in the presence or absence of FGF-2 and its cofactor heparin. Experiments were performed in triplicate twice and LECs with a passage number 2 were used. siCTR, control siRNA for luciferase; siFGFR3, siRNAs for FGFR-3; *, p<0.05; **, p < 0.01; ***, p < 0.001.
3.1.2.6 FGF-2 binds directly to low and high affinity receptors in LECs and subsequently internalized for degradation
We next investigated whether FGF ligands physically interact with FGF receptors present in lymphatic endothelial cells. LECs were incubated with increasing amounts of 125I-FGF-2 and levels of binding to the low and high affinity sites were determined.
48 125I-FGF-2 was bound in a concentration-dependent manner to LECs but binding was not fully saturable for the low affinity binding sites (Fig. 3.1.5A). For low affinity binding (proteoglycans), a Kd of 1 nM and 400,000 binding sites/cell were determined (Fig. 3.1.5A). For high affinity binding sites (receptors), maximum binding was detected between 4-6 ng/ml of 125I-FGF-2 (Fig. 3.1.5B). Scatchard analysis revealed high affinity binding (Kd of 72 pM) and approximately 5,300 binding sites per cell (Fig. 3.1.5B). These values are similar to those found on vascular endothelial cells (Moscatelli, 1987). We then determined internalization of 125I-FGF-2 in LECs. Between 1 to 4 hours, the internalization rate was 0.046 ng/h/105 cells (Fig. 3.1.5C). This value progressively decreased between 4-8 hours (0.01 ng/h/105 cells) and 8-12 hours (0.004 ng/h/105 cells), indicating that FGF-2 internalization progressively slows down with time. After 1 hour of internalization, a fragment of 15-kDa (together with the 18-kDa band) was detected (Fig. 3.1.5D). At 1.5 and 2 hours, two additional fragments of 10- and 8-kDa appeared and their amounts increased with time. Maximum degradation was observed between 12-24 hours. Interestingly, at 24 hours, 18-kDa FGF-2 was still present in significant amounts in LECs. Taken together, our biochemical study provides a detailed information on binding kinetics of FGF-2 to its receptors and subsequent internalization and degradation patterns of the ligand in lymphatic endothelial cells, which are largely comparable with those of vascular endothelial cells as previously described (Bikfalvi et al, 1989).
49 A B 0.020
~ "'0 0.12 c- 0.015 c~ ,!!! , . 0- 0= ~ ~ ~ . N "'u 0.08 ryö '0 0.010 ~ ~- ~~ iri;u,••• ~ ~ S .s 0.04 ; '::EJ.- ; 0.005 .- . • . ~""'" 0._ 0.0'. 0,0'>0 0.00 ~---_---_---,-=~,~~.~-= _1""0',_) o , 4 6 0.000 0 , 4 6 12llI·FGF·2 (ng/ml) 1251_FGF_2 (nglmlj
c 0.3$ D 0.5 1 1.5 2 4 8 12 24 (hr) • • -18 kDa • -15 kDa - JO kDa - 8 kDa o.ooli--_--_--_-_--~ o 10 1S ~ ZS Time(h)
Figure 3.1.5 Binding, internalization and degradation of FGF-2 in LECs. Concentration dependency of 125I-FGF-2 binding to low affinity sites (A) and high affinity receptors (B). Cells were incubated with increasing concentrations of 125I-FGF-2 and the specific binding was determined as described in Materials and Methods. Scatchard plots are shown in insets. (C) Internalization of 125I- FGF-2 was determined by incubating cells with 10 ng/ml 125I-FGF-2 at 37 °C for specified time intervals. (D) After internalization, solubilized cell extracts were run on a 15 % SDS-PAGE gel, dried and processed for autoradiography (PhosphorImager) to visualize the degradation profile. Time (hours) after incubation and molecular mass of the degraded products are shown. The data are representative for two independent experiments performed in duplicates.
3.1.2.7 FGF signaling regulates migration, proliferation and apoptosis of cultured primary lymphatic endothelial cells
We next investigated the effects of two specific FGF ligands on migration, proliferation and apoptosis of primary human LECs. Treatment with recombinant human FGF-1 and FGF-2 significantly enhanced migration and proliferation of LECs (Fig. 3.1.6A and B). Furthermore, both FGF ligands protected LECs from apoptosis induced by serum depletion (Fig. 3.1.6C). A previous in vivo study in mouse corneas indicated that FGF-2 might indirectly promote lymphangiogenesis through activation of the VEGF-C/VEGFR-3 pathway (Kubo et al, 2002). To determine if FGF-2 can stimulate LEC migration in vitro directly or indirectly, we studied the effect of FGF-2 in the presence or absence of an anti-VEGFR-3 blocking antibody. Both VEGF-C and
50 FGF-2 stimulated the migration of LECs at a comparable level (Fig. 3.1.6D). However, neutralization of VEGFR-3 abrogated the enhanced migration of LECs by VEGF-C, but not by FGF-2, indicating that FGF-2 can function independently of the VEGF-C/VEGFR-3 pathway in vitro.
A Migration B Proliferation 5000 700 4500 600 4000 *** ** ,..:Vl *** ~ c: 500 'e 3500 :::l :3 Cl) Q) 3000 .'.~. U 400 u c: ~ 2500 Cl) u u Vl 300 ~ 2000 ~ o o if 1500 :::l 200 LL. 1000 100 500 o o
C 120- Apoptosis D Migration 700 ** :ß '00 E o 600 ** l3 I *** Q) 80 ~ 11 u c: 500 :3 C :::l I o ** * Cl) ,2> 60 CJ 400 c: o Cl) u :;l 300 'E Cl) 1Il 40 ... o Ci'" :::l 200 o u:: ~ 20 100
o o
Figure 3.1.6 Stimulatory effects of fibroblast growth factors on proliferation, migration and survival of LECs. (A) Migration of LECs was promoted by FGF-1 and FGF-2. Cells were allowed to migrate toward fibronectin in serum-free media containing FGF-1 or FGF-2 (10 ng/ml) in the presence of heparin (1 ug/ml). Numbers of migrated cells were quantified by fluorescence assay. (B) FGF-1 and FGF-2 stimulated LEC proliferation. LECs were treated with or without FGFs for 48 hours. Increase in cell numbers was determined using the MUH fluorescence assay. (C) FGF-1 and -2 (10 ng/ml) inhibit LEC apoptosis induced by serum depletion for 24h. Addition of 20% serum, but not of heparin alone, prevented LEC apoptosis. Data are expressed in % of BSA control and are shown as means±SD. (D) FGF-2 directly promoted LEC migration independently from VEGFR-3 activation. LEC migration was stimulated by VEGF-C (100 ng/ml) or FGF-2 (10 ng/ml), but the enhanced migration by VEGF-C was abrogated by addition of an anti-VEGFR-3 blocking antibody. *p<0.05, **p<0.01, ***p<0.001.
51 3.1.3 Discussion
The homeodomain protein Prox1 plays an essential role in the lymphatic system development during embryogenesis as a master regulator that induces lymphatic lineage-specific differentiation (Hong & Detmar, 2003b; Hong et al, 2002; Petrova et al, 2002b; Wigle et al, 2002b; Wigle & Oliver, 1999). Furthermore, the LEC lineage- specification occurring during embryogenesis can be post-developmentally recapitulated when Prox1 is ectopically expressed in neonatal BECs (Hong & Detmar, 2003b; Hong et al, 2002; Petrova et al, 2002b). However, the molecular mechanisms underlying the cell fate decision controlled by Prox1 remained to be studied. In this report, we identified FGFR-3 as an initial Prox1 target gene during the early lymphatic system development. This upregulation is mediated at the transcriptional level by a direct binding of Prox1 to the specific sequence elements in the FGFR-3 promoter. Consistently, FGFR-3 is strongly expressed in the vein-derived lymphatically differentiating endothelial cells and in post-developmental lymphatic vessels in neonatal human foreskins. We also found that FGFR-3 plays an important role in mediating proliferating signals of LECs. Furthermore, our biochemical study demonstrated that FGF-2 bind to the low and high affinity receptors in LEC to promote migration, proliferation and cell survival of LECs independently of the VEGF-C/VEGFR-3 signal pathway.
Lymphatic endothelial cells are derived from venous endothelial cells that are of mesodermal origin. Our finding that Prox1 specifically upregulates the IIIc variant of FGFR-3, the major isoform in LEC, is consistent with previous studies that the IIIc forms of FGF receptors (FGFR-1 to -3) are mainly expressed by the mesenchymal lineage cells (Alarid et al, 1994; Orr-Urtreger et al, 1993; Yan et al, 1993). Interestingly, FGF receptors and their splicing variants exhibit strikingly distinct binding affinities to different FGF ligands (Ornitz & Itoh, 2001; Ornitz et al, 1992; Powers et al, 2000). As an example, the FGFR-3 IIIb isoform interacts with FGF-1, but not with FGF-2, FGF-4, or FGF-6, whereas the IIIc isoform is activated by all of these ligands to promote fibroblast proliferation (Kanai et al, 1997; Ornitz et al, 1996). Furthermore, FGFR-3 IIIc also displays a high affinity to FGF-8, FGF-17 and FGF-18 (Liu et al, 2002; Xu et al, 1999; Xu et al, 2000). Given these facts and our
52 findings presented here, upregulation of FGFR-3 IIIc by Prox1 in the LEC-specific fashion may be essential for mediating proliferation signals for the lymphatic system development, which may be distinct from signals for the blood vascular system development. This notion of differential proliferation signal is highly conceivable because only a subset of endothelial cells in the developing vein needs to be activated to proliferate and migrate out to form initial lymphatic vessels during embryogenesis. Therefore, FGFR-3 may be one of the major players in the molecular mechanism responsible for the LEC differentiation and subsequent lymphatic system development. Furthermore, the expression and maintenance of an additional FGF receptor may be also advantageous for the function of the lymphatic system. Because the lymphatic system plays essential roles in various aspects of the immune system, FGFR-3 may be important for cross talks between LECs and immune cells. It will be interesting to study the role of FGFR-3 during tissue repair, inflammation, and tumor development and metastasis.
We found that interaction of Prox1 with a specific DNA sequence element in the FGFR-3 promoter was necessary for the Prox1-mediated transcriptional activation of FGFR-3. The Prox1-binding sequences found in the FGFR-3 promoter consist of two overlapping consensus binding sequences of Prospero (Cook et al, 2003; Hassan et al, 1997). These sequence motifs, conserved between the mouse and human FGFR-3 genes, form a complex with purified GST-Prox1 protein and were sufficient to re- instate the Prox1-mediated transcriptional activation to a non-activating reporter vector. Previously, functional interactions of Prox1 with other transcriptional regulators were reported in the developing lens. The sequence-specific Six3 repressor antagonizes the Prox1 activation of the γ–crystallin promoter (Lengler et al, 2001). Similarly, Pax-6 occupies a specific sequence motif and prevents Prox1-mediated activation of the βB1-crystallin gene in chicken lens epithelial cells, whereas Prox1 binds to the same site to activate the gene in lens fiber cells (Cui et al, 2004). In contrast, Prox1 was shown to function as a corepressor of Ff1b, the Zebra fish homologue of mammalian steroidogenic factor-1 (SF-1) by a direct protein-protein interaction during embryonic development of the interrenal primordium (Liu et al, 2003b). It remains unknown if these Prox1 interacting partners also play a role in the development of the lymphatic system. Because Prox1 activates some genes but
53 represses others in lymphatically differentiating endothelial cells, it will be important to characterize transcriptional factors involved in this regulation during lymphatic development.
The VEGF-C/VEGFR-3 signaling was shown to play an essential role in the development of the lymphatic system (Karkkainen et al, 2004). Promotion of lymphangiogenesis by FGF-2 in mouse corneas was suggested to be mediated through upregulation of VEGF-C by stromal cells and FGF-2-induced corneal lymphangiogenesis was abrogated by a neutralizing antibody against VEGFR-3, the major receptor for VEGF-C (Chang et al, 2004; Kubo et al, 2002). In contrast, we found specific expression of FGFR-3 in LECs in vitro and in vivo and direct binding of FGF-2 to low- and high affinity receptors in LECs. In addition, we found that FGF- 1 and FGF-2 can enhance migration, proliferation and survival of LECs and that the FGF-2-mediated activation of LEC migration is not dependent on the function of VEGFR-3. These results clearly indicate that these FGF ligands directly bind to their receptors in LEC and exert a direct role in lymphatic vessel formation. Nonetheless, our data do not rule out an indirect activation of FGF ligands through VEGFR-3 because our experiments involved only purified LECs, but not accompanying other stromal cells, the proposed source of VEGF-C (Chang et al, 2004; Kubo et al, 2002). Therefore, FGF ligands may exert their functions in multiple manners depending on the tissue microenvironment. Our finding that LECs expressed an additional FGF receptor is of particular interest because a recent study showed that lymphangiogenesis occurred at a low dosage of FGF-2 (12.5 ng), a concentration that did not induce accompanying angiogenesis in the mouse cornea assay (Chang et al, 2004). Therefore, it is conceivable that LECs may be more sensitive to FGF-2 stimulation than BECs due to expression of additional FGF receptors.
FGFR-3 has been previously shown to be essential for various developmental processes such as bone morphogenesis, inner ear development and alveogenesis in the lung (Ornitz & Marie, 2002; Weinstein et al, 1998). Because we found that FGFR-3 is a target gene of Prox1 and that Prox1 specifies lymphatic endothelial cell fate, we investigated if FGFR-3 mediates an inductive signal for lymphatic differentiation and found that knockdown of FGFR-3 mRNA significantly inhibited LEC proliferation.
54 This suggests that the receptor may play an important role in mediating cell proliferation during lymphatic system development. Our preliminary study indicates that the FGFR-3 null mice developed apparently normal lymphatic capillaries in the skin. We believe that this is most likely due to functional complementation by other FGF receptors. This notion of functional cooperation among FGF receptors is further supported by a study of the FGFR-3 and FGFR-4 double knockout mice (Weinstein et al, 1998). Homozygous fgfr-3-/-fgfr-4-/- mutant mice displayed abnormal alveogenesis during lung development, a phenotype that was not present in single knockout mutants, suggesting that the two FGF receptors function together to direct normal lung development. It will be of great interest to evaluate lymphatic vessel development in the fgfr-3-/-fgfr-4-/- mutant mice. Furthermore, mice lacking FGF-18 display a similar mutant phenotype in bone morphogenesis as FGFR-3 null mice, defining FGF-18 as a physiological ligand for FGFR-3 during bone development (Liu et al, 2002). It will be also interesting to see whether FGF-18 single or FGFR-3/FGF- 18 double knockout mice develop a normally functioning lymphatic system.
55 3.2 Quantification of vascular lineage-specific differentiation and molecular characterization of in vivo (lymph)angiogenesis by a novel low-density microvascular differentiation array
3.2.1 Introduction
The formation and activation of blood vascular and lymphatic endothelium have an important role in the progression and metastasis of the majority of human cancers (Alitalo et al, 2005; Carmeliet, 2003). Tumors need to induce the growth of new blood vessels (angiogenesis) in order to secure the sufficient supply of oxygen and nutrients, and the growth of new lymphatic vessels (lymphangiogenesis) has been shown to promote cancer metastasis to sentinel lymph nodes and beyond (Hirakawa et al, 2006; Hirakawa et al, 2005b; Mandriota et al, 2001; Skobe et al, 2001; Stacker et al, 2001). Recent studies indicate that both types of endothelium are also involved in chronic inflammatory diseases such as rheumatoid arthritis, inflammatory bowel disease and psoriasis (Alitalo et al, 2005; Carmeliet, 2003; Cueni & Detmar, 2006b). As a result, there has been a surge of interest in identifying novel targets that can be used to specifically image these processes and to target them therapeutically. However, these types of studies have been hampered by the lack of identified lymphatic-specific markers and growth factors.
The lymphatic system is a unidirectional vascular network that drains fluids and cells from peripheral tissues and attracts and transports antigen-presenting cells to mediate the afferent immune response (Oliver & Detmar, 2002). During embryonic development, lymphatic progenitor cells bud off from embryonic veins under the influence of the transcription factor Prox1, migrate, form lymph sacs and eventually form mature lymphatic vessels (Alitalo et al, 2005; Oliver, 2004). Lymphatic endothelial cells (LEC) and blood vascular endothelial cells (BEC) therefore share a large number of common endothelial lineage genes, and there are only a few specific known marker genes, such as the hyaluronan receptor LYVE-1 (Prevo et al, 2001) and the mucin-type glycoprotein podoplanin (Schacht et al, 2003), that distinguish lymphatic vessels from blood vessels (Cueni & Detmar, 2006b).
56 There have been previous attempts to identify the lineage-specific transcriptomes of LEC and BEC using gene microarrays (Hirakawa et al, 2003; Petrova et al, 2002a; Podgrabinska et al, 2002). However, the arrays used in these analyses included an incomplete set of human genes and a large-scale confirmation of the results by other methods has not been attempted. Gene expression profiling is a time-consuming, relatively expensive process that requires specialized equipment, so it is not practical for use in characterizing the (lymph)angiogenic activity of tissues samples. A simple and rapid assay for the quantitative analysis of in vitro vascular differentiation, and of angiogenesis and lymphangiogenesis in tissue samples would provide a major technological advance for research and clinical analysis.
We aimed to comprehensively identify the lineage-specific transcriptomes of primary human LEC and BEC. In this study, we used the Applied Biosystems Human Genome Survey v.2 (AB-HGS), which includes almost all of the known human genes, to identify lymphatic and blood vascular signature genes.
Using the LEC and BEC genes identified, we developed a novel, TaqMan RT-PCR based low-density microvascular differentiation array (LD-MDA) in a microfluidic card format to allow for the simultaneous quantification of 96 genes. Using the LD- MDA we were able to reproducibly identify and quantify the differentiation of LEC and BEC cells in vitro based on expression levels of the genes analyzed in the assay. We then designed and developed a computational algorithm to systematically identify genes associated with (lymph)angiogenic activity in tissues samples obtained from patients with the chronic inflammatory skin disease.
3.2.2 Results
3.2.2.1 Comprehensive identification of vascular lineage-specific gene signatures
We first analyzed the transcriptional profiles of three matched pairs of LEC and BEC using the AB-HGS microarrays. Genes that were expressed at ≥ 2-fold higher levels in LEC than in BEC (and vice versa) in all three independent pairs of LEC-BEC were
57 considered to be endothelial lineage-specific signature genes. Based on these criteria we identified a total of 236 LEC signature genes (upregulated ≥ 2-fold in LEC) and
342 BEC signature genes (upregulated ≥ 2-fold in BEC) (Appendix Table 1). Genes that were highly expressed specifically by LEC included previously identified LEC- associated genes such as Prox1, podoplanin, carcinoembryonic antigen-related cell adhesion molecule-1 (CEACAM1) and soluble guanylate cyclase 1 alpha 3, and genes whose expression was not before associated with LEC, including dipeptidyl peptidase IV (DPPIV) and collectin 12 (COLEC12) (Table 3.2-1a). We also identified several BEC signature genes that had not been previously associated with this cell type, including urokinase plasminogen activator (PLAU), membrane metallo-endopeptidase (MME) and endothelial lipase (LIPG) (Table 3.2-1b). Together, these results establish a more comprehensive catalogue of endothelial lineage-specific gene signatures.
58 Table 3.2-1a Top 40 LEC signature genes in all three matched-pair samples by microarray (sorted by median)
LEC gene Fold change signature AB probe Sample Sample Sample Symbol Gene name ID 1 2 3 GUCY1A3 guanylate cyclase 1, soluble, alpha 3 170165 497.26 154.95 83.6 GUP1 GRINL1A complex upstream protein 104996 193.23 13.75 9.04 HS3ST1 heparan sulfate (glucosamine) 3-O-sulfotransferase 1 154628 189.98 9.13 2.51 PDK4 pyruvate dehydrogenase kinase, isoenzyme 4 101060 175.09 12.45 14.72 CH25H cholesterol 25-hydroxylase 117883 130.44 3.56 15.25 MRC1 mannose receptor, C type 1 198568 119.37 6.39 15.29 GIMAP5 GTPase, IMAP family member 5 177981 97.71 6.23 5.6 EDNRB endothelin receptor type B 150558 83.94 6.56 9.16 HYAL1 hyaluronoglucosaminidase 1 184118 75.23 5.75 7.2 RBP1 retinol binding protein 1, cellular 149921 69.97 13.94 7.51 C6orf123 chromosome 6 open reading frame 123 105756 69.4 10.64 7.36 C2orf23 chromosome 2 open reading frame 23 156624 68.65 13.48 7.75 DKFZP586A0522 DKFZP586A0522 protein 107957 64.69 4.03 12.63 ST6GALNAC3 ST6...N-acetylgalactosaminide alpha-2,6-sialyltransferase 3 189728 55.54 3.78 4.99 CEACAM1 carcinoembryonic antigen-related cell adhesion molecule 1 219223 50.13 11.73 3.9 CD36 CD36 antigen (collagen type I receptor, thrombospondin receptor) 121773 49.85 4.44 10.57 DNASE1L3 deoxyribonuclease I-like 3 167226 49.27 2.55 51.27 SEPP1 selenoprotein P, plasma, 1 169984 48.97 6.42 8.18 IQCA IQ motif containing with AAA domain 152027 46.85 11.54 4.4 CETP cholesteryl ester transfer protein, plasma 140569 45.16 3.66 2.84 TFF3 trefoil factor 3 (intestinal) 114445 44.84 8.11 12.68 ADAMTSL3 ADAMTS-like 3 158085 43.95 2.18 25.43 XLKD1 extracellular link domain containing 1 195865 43 0.98 2.01 RBM35B RNA binding motif protein 35B 167987 41.06 10.21 14.62 TMEM88 transmembrane protein 88 200951 40.58 3.49 5.61 COLEC12 collectin sub-family member 12 114422 39.34 17.2 6.12 CYP1A1 cytochrome P450, family 1, subfamily A, polypeptide 1 135086 39.01 3.21 6.51 PROX1 prospero-related homeobox 1 124383 38.93 6.77 9.56 PPARG peroxisome proliferative activated receptor, gamma 192239 37.79 60.22 10.05 ZNF467 zinc finger protein 467 184463 36.77 8.26 2.84 GMFG glia maturation factor, gamma 180184 36.59 2.91 7.12 DPP4 dipeptidylpeptidase 4 (CD26) 209451 35.86 21.59 12.1 ABCA4 ATP-binding cassette, sub-family A (ABC1), member 4 194955 35.51 26.3 7.15 IL7 interleukin 7 127208 34.58 2.55 17.78 PCSK6 proprotein convertase subtilisin/kexin type 6 154864 32.87 3.79 4.71 TRPC6 transient receptor potential cation channel, subfamily C, member 6 101144 32.12 3.07 9.43 PDPN podoplanin 219722 30.81 2.48 5 C17orf28 chromosome 17 open reading frame 28 115291 29.54 5.41 2.52 MAF v-maf musculoaponeurotic fibrosarcoma oncogene homolog 186589 28.43 10.69 4.2 C18orf30 chromosome 18 open reading frame 30 171508 28.33 2.77 4.66
59 Table 3.2-1b Top 40 BEC signature genes in all three matched-pair samples by microarray (sorted by median)
BVEC gene Fold change signature AB probe Sample Sample Sample Symbol Gene name ID 1 2 3 COL6A3 collagen, type VI, alpha 3 115643 884.5 44.24 2.82 ADAMTS1 a disintegrin-like and metalloprotease with thrombospondin type 1 216353 477.78 45.43 4.4 motif, 1 COL1A2 collagen, type I, alpha 2 105493 344.17 86.94 3.29 CRISPLD2 cysteine-rich secretory protein LCCL domain containing 2 170538 331.32 29.54 5.35 BEX1 brain expressed, X-linked 1 137034 250.21 88.31 7.02 GPNMB glycoprotein (transmembrane) nmb 161212 236.81 20.27 4.14 PTGFR prostaglandin F receptor (FP) 103022 207.64 6.58 5.69 PLAU plasminogen activator, urokinase 208672 165.63 6.01 126.53 CDH2 cadherin 2, type 1, N-cadherin (neuronal) 187321 163.67 5.19 17.45 NRG1 neuregulin 1 223108 159.16 13.42 8.43 AMIGO2 amphoterin induced gene 2 154434 157 7.48 3.03 GFPT2 glutamine-fructose-6-phosphate transaminase 2 113797 155.06 66.02 3.64 OXTR oxytocin receptor 200205 155.01 23.39 3.5 FAP fibroblast activation protein, alpha 164725 131.07 47.35 46.66 GLIPR1 GLI pathogenesis-related 1 (glioma) 117689 118.71 32.66 9.84 MME membrane metallo-endopeptidase (CALLA, CD10) 197353 117.83 4.74 9.82 CSPG2 chondroitin sulfate proteoglycan 2 (versican) 207524 117.52 21.87 4.79 SYTL2 synaptotagmin-like 2 118410 115.03 22.51 17.59 LOXL1 lysyl oxidase-like 1 156579 102.71 8.15 17.68 PCSK1 proprotein convertase subtilisin/kexin type 1 213177 96.18 8.44 91.13 RGS4 regulator of G-protein signalling 4 165955 93.55 15.49 10.88 FLT1 fms-related tyrosine kinase 1 219494 93.14 21.3 35.73 C7orf10 chromosome 7 open reading frame 10 180432 86.96 15.46 13.27 SHRM shroom 207317 84.74 10.08 3.83 LOC152573 hypothetical protein BC012029 150646 70.98 8.95 34.85 BASP1 brain abundant, membrane attached signal protein 1 198318 69.8 7.59 25.75 COL5A1 collagen, type V, alpha 1 110570 68.34 5.66 30.97 EMP3 epithelial membrane protein 3 152376 64.94 11.72 14.52 COL6A1 collagen, type VI, alpha 1 215580 63.74 4.99 4 IL1RL1 interleukin 1 receptor-like 1 131513 62.75 36.34 24.5 VEGFC vascular endothelial growth factor C 170337 59.63 10.14 5.56 LIPG lipase, endothelial 200619 59.54 112.03 50.64 LCP1 lymphocyte cytosolic protein 1 (L-plastin) 175091 57.18 4.02 12.13 TAGLN transgelin 172572 56.92 6.98 47.13 NUDT11 nudix (nucleoside diphosphate linked moiety X)-type motif 11 125359 56.13 7.6 9.85 FAM20C family with sequence similarity 20, member C 199772 54.95 8.38 2.06 FAT FAT tumor suppressor homolog 1 (Drosophila) 131558 54.78 54.6 8.65 IL7R interleukin 7 receptor 200834 54.27 7.7 2.93 TCEAL7 transcription elongation factor A (SII)-like 7 130055 53.18 9.41 17.62
3.2.2.2 Identification of lineage-specific biological functions by in silico analysis
We next investigated whether the establishment of comprehensive LEC and BEC gene signatures could be used to identify lineage-specific biological functions for each cell type using in silico molecular pathway analysis. We found that genes involved in fatty acid, cholesterol and steroid metabolism were significantly over-
60 represented among LEC signature genes (P < 0.0005), as compared with BEC signature genes (Table 3.2-2). In contrast, genes involved in angiogenesis and blood clotting were significantly over-represented among the BEC signature genes (P < 0.005) but not among the LEC signature genes. Genes involved in cell adhesion, immunity and defense, and cell structure and motility were overrepresented in both the LEC and BEC signatures (P < 0.05), indicating common biological functions of both endothelial cell types (Table 3.2-2).
61 Table 3.2-2 Biological process analysis of LEC and BEC signature genes using the Panther Classification System
LEC BVEC Biological Process genes expected p-value genes expected p-value (226) (337) Carbohydrate metabolism 15 4.5 *** 7 6.72 ns Cell cycle control 12 3.06 *** 6 4.57 ns Lipid, fatty acid and steroid metabolism 17 6.04 *** 10 9.01 ns
mRNA transcription 27 13.76 *** 19 20.51 ns Nucleoside, nucleotide and nucleic acid 40 24.68 ** 31 36.8 ns process
specific biological metabolism - Embryogenesis 5 1.04 ** 3 1.56 ns LEC Neuronal activities 11 4.35 ** 10 6.49 ns
Cell adhesion-mediated signaling 6 2.76 ns 22 4.12 *** Skeletal development 1 0.95 ns 12 1.42 *** Proteolysis 11 7.1 ns 25 10.59 *** Extracellular matrix protein-mediated 0 0.51 ns 6 0.76 *** signaling process
specificbiological Angiogenesis - 2 0.42 ns 5 0.63 *** Blood clotting 1 0.68 ns 5 1.02 **
BVEC MAPKKK cascade 4 1.41 ns 7 2.1 **
Signal transduction 63 25.45 *** 92 37.95 *** Developmental processes 48 15.99 *** 61 23.84 ***
Neurogenesis 16 4.35 *** 18 6.48 *** Cell proliferation and differentiation 22 8.01 *** 26 11.94 *** Cell communication 23 8.9 *** 52 13.28 *** Protein modification 20 8.75 *** 22 13.04 * Oncogenesis 11 3.51 *** 20 5.23 *** Receptor protein tyrosine kinase signaling 7 1.61 ** 6 2.4 * pathway specific biologicalprocess - Cell adhesion 12 4.28 ** 31 6.39 *** Protein phosphorylation 13 5.09 ** 17 7.59 ** Tumor suppressor 4 0.69 ** 8 1.03 *** Cell structure and motility 18 9.32 ** 30 13.89 ***
Endothelialcell Cytokine and chemokine mediated signaling 6 1.93 * 7 2.88 * pathway Immunity and defense 19 11.34 * 34 16.91 ***
*** p-value < 0.0005, ** p-value < 0.005, * p-value < 0.05, ns = not significant
62 3.2.2.3 Using the LD-MDA to quantify lineage-specific endothelial cell differentiation
We next aimed to develop a simple and rapid assay to quantify the degree of lineage- specific differentiation in human endothelial cell samples. We selected 54 genes from the LEC signature and 31 genes from the BEC signature, based upon the degree of specificity determined by array analysis and their potential function (Appendix Table 1), and 5 pan-endothelial genes, which are strongly expressed in both cell types (PECAM-1, vWF, KDR, Tie2 and CDH5) as general markers of endothelial lineage differentiation. The expression of these genes and of 6 housekeeping genes were quantified in the TaqMan-based low-density LD-MDA using 384-well microfluidic cards. Using this array, the differential expression levels of all 85 LEC and BEC signature genes were confirmed using mRNA from same three matched pairs of LEC and BEC that were used for the gene array studies (Appendix Table 1).
We next investigated whether the LD-MDA might be used to quantify the level of differentiation among different endothelial cell types. The LD-MDA was performed on 10 primary human dermal LEC and 8 human dermal BEC samples previously isolated in our laboratory. Additional samples included human dermal microvascular endothelial cells (HDMEC; n = 2), human umbilical vein endothelial cells (HUVEC; n = 2), the immortalized HDMEC cell line HMEC-1, HaCaT keratinocytes and dermal fibroblasts. We established an ‘endothelial lineage score’ (ELS) by subtracting the sum of the normalized cycle values (ΔCt) of all 54 LEC-specific genes from the sum of the ΔCt values of all 31 BEC-specific genes. As a measure for the degree of endothelial cell differentiation, the sum of the ΔCt values of the five pan-endothelial marker genes was calculated for each sample and defined as vascular lineage score (VLS). This analysis revealed that all LEC samples clustered together, with ELS scores ranging from 98 to 442, whereas the BEC samples had ELS scores ranging from 825 to 1107 (Fig. 3.2.1A). ELS scores for HUVEC fell into the BEC range (1089 and 1081, respectively), in agreement with the blood vascular origin of HUVEC. In contrast, the HMEC-1 cell line had a higher VLS score (96) than the BEC (32-57). This was likely because of the lower levels of expression of the pan-
63 endothelial markers VE-cadherin, VEGFR-2 and vonWillebrand factor in the HMEC- 1 cell line. The two non-endothelial cell types (keratinocytes and dermal fibroblasts) had ELS scores of 987 and 1317 and VLS scores of 179 and 195, respectively, clearly discriminating them from the endothelial cell lines tested in the LD-MDA. A 250
Fibroblasts 200 HaCaT _
150 III - ...J HMEC-l > 100 ... LEC BVEC 0 50 ••• ••• o:B=Oo -HUVEC 0 0 200 400 600 800 1000 1200 1400 ELS B e 100 R'= 0.9902 :::l .--- X ---Ü 80 HDMEC --~ 'E (84.1% LEC) - .= 60 () W ...J 40 ...0 ~ 20 -.
200 360 600 660 800 ELS C 18.3 79.8 105
104 ...... , C U 103
102 0 -;.,;,n;r.....I-~rTT'.",,~-rrm,.,.-rTTT'~"19 105
Podoplanin
Figure 3.2.1 Quantification of lineage-specific gene expression and differentiation of endothelial and non-endothelial cell types using LD-MDA. (A) LD-MDAs were performed on total RNA obtained from 10 different primary human LEC (black circles), from 8 BEC (white circles), from 2 HUVEC (grey circles), from HMEC-1 (grey triangle), from HaCaT keratinocytes and from dermal fibroblasts (black squares). Determination of the endothelial lineage score (ELS) and of the vascular lineage score (VLS) revealed that all LEC samples clustered together, whereas the BEC samples were clearly distinguished by their higher ELS scores. ELS scores for HUVEC were indistinguishable from BEC, whereas immortalized HMEC-1 cells did not cluster with BEC, because of their higher VLS score. Keratinocytes and fibroblasts were clearly discriminated from all endothelial cells, based on their VLS scores. (B) Analysis of defined mixtures of LEC and HUVEC by LD-MDA revealed that the distribution of the samples (mixtures) on the ELS-axis showed a linear correlation to the percentage of LEC in each sample (R2 = 0.9902). Using the equation Y (% of LEC) = -1.0865x+3.9474, the percentage of LEC in the examined HDMEC was predicted as 84.1 %. (C) Flow cytometry analysis of the same HDMEC sample stained for the panendothelial marker CD31 and for the LEC-specific marker podoplanin revealed that HDMEC contained approximately 80% LEC.
64
To further test this technology, defined mixtures of LEC and HUVEC (percent of LEC: 100, 80, 50, 20, and 0) were analyzed by LD-MDA. The distribution of the samples (mixtures) on the ELS score was indeed proportionally correlated to the percentage of LEC in each sample (Fig. 3.2.1B). To investigate whether the LD- MDA could also be used to quantify the percentage of LEC in HDMEC cultures, which represent a mixture of LEC and BEC, we used LD-MDA data from the mixtures of LEC and HUVEC as a standard to derive the linear relationship between ELS and the percentage of LEC in the mixture. A strong linear relationship was identified with R2 = 0.9902 and the equation of Y (% of LEC) = -1.0865x + 3.9474 (Fig. 3.2.1B). Using this equation, we were able to predict the percentage of LEC in a representative HDMEC culture as 84.1 %. Importantly, this prediction - based on the LD-MDA data that were derived from mRNA expression levels - was confirmed at the level of protein expression because FACS analysis revealed that approximately 80% of HDMEC were CD31-positive/podoplanin-positive LEC (Fig. 3.2.1C). Together, these findings indicate that the LD-MDA is a new tool for quantifying the degree of endothelial lineage-specific differentiation of cultured cells and for determining the purity of endothelial cell cultures.
3.2.2.4 Hierarchical clustering according to endothelial lineage-specific gene signatures
We next asked whether a subset of genes might be identified that shows the most consistent lineage-specific expression. Thirty-nine out of 95 genes selected for the LD-MDA had little variation in expression levels between different samples of the same cell type, but large differences in expression levels between LEC and BEC (Appendix Table 1). Based on the expression levels of these core differentiation genes, hierarchical clustering separated the different cell types into four distinct clusters: LEC, BEC, HMEC-1 and non-endothelial cells (Fig. 3.2.2). We also identified four groups of genes that were expressed at high levels in one cell type (LEC versus BEC) and at moderate or low levels in another cell type, or that were expressed at moderate levels but were not expressed at all by the other cell type.
65 nodes 1 4 12 23 53.6 766 100 similarity (%) LEC-7
"Ct 19
10.5
L------IHMEC-1
,------1HeC8t L- I_
Figure 3.2.2 Hierarchical clustering according to endothelial lineage-specific gene signatures. Hierarchical clustering, based on the expression levels of 39 core differentiation genes, separated the different cell types into four distinct clusters: LEC, BEC, HMEC-1 and non-endothelial cells. HUVEC cells were separated slightly from the BEC cluster. Two groups of genes were identified for each of the LEC/BEC pairs that were either expressed at high levels in one cell type and at moderate−low levels in the other cell type ("highly expressed genes") or expressed at moderate levels in one cell type and not expressed by the other cell type ("specific genes").
3.2.2.5 Identification of (lymph)angiogenic-mediators using a novel Prediction Relevance Ranking analysis
We next investigated whether the LD-MDA could be used to identify key endothelial signature genes associated with in vivo inflammation by quantitatively profiling 43 samples obtained from psoriatic skin lesions. Psoriasis is a chronic inflammatory skin disease with prominent angiogenesis and lymphangiogenesis(Kunstfeld et al, 2004). Half of the skin samples were subjected to differential immunofluorescence analyses by staining for the lymphatic marker LYVE-1 and for the vascular marker CD31 (Fig. 3.2.3A-D). The computer-assisted morphologic analysis revealed different degrees of expansion of the dermal lymphatic and blood vascular networks amongst psoriatic lesions, as evaluated by the relative tissue area occupied by lymphatic or blood vessels (Fig. 3.2.3E). RNA was isolated from the other half of each sample and was subjected to expression quantification by LD-MDA. We used a Prediction Relevance
66 Ranking (PRR) analysis to learn all possible models up to degree four:
. PRR characterized the whole model space and provided a ranking of the most predictive genes for each of the two targets: lymphatic vessel area (LVA) and blood vessel area (BVA) (Fig. 3.2.3F). In addition to PRR analysis, we partitioned the samples into four groups by using the median of LVA and BVA as thresholds. We then applied analysis of variance (ANOVA) to assess the influence of specific genes within these groups. Comparing the group with low LVA and low BVA (Fig. 3.2.3E; quadrant IV) against the group with high BVA (Fig. 3.2.3E; quadrants II+III) revealed a potential involvement of FLT1, FGF12, ADD3, ALDH1A1, MRC1 and IL7R in inflammatory angiogenesis (Fig. 3.2.3F). Furthermore, comparing the group with low LVA and low BVA (Fig. 3.2.3E; quadrant IV) against the group with high LVA (Fig. 3.2.3E; quadrants I+II) revealed involvement of FLT1, FGF12, IL7R, ADD3, MRC1, INHBA, CDH11, LMO2, RELN and KDR in inflammatory lymphangiogenesis (Fig. 3.2.3F). Comparing PRR and ANOVA results revealed FGF12, IL7R and FLT1 as the most significant factors involved. Since FGF12 or IL7 had not been previously implicated in (lymph)angiogenesis, we next treated cultured human LEC and BEC with FGF12 or IL7 and investigated their effects on cellular proliferation. We found that both IL7 and FGF12 significantly induced the proliferation of LEC and of BEC (Fig. 3.2.3G, H).
67 E I 11
.. ... --.. ... • A.
BVA(%)
F Q 1+11 (LV) Q 1I+IlI(BV) PRR Rank LVA eVA ANOVA ANOVA vs Q IV vs Q IV 1 INHBA FLT1 FLT1 0.00115 FLT1 0.00340 ~2 DPP4 ~~ FGF12 000314 ....l'.GF12 000594 3 tL7R ITGBJ AOO3 000862 AOD3 0.03561 4 TEK FG~ MRC1 001503 ALDH1A1 004141 5 ADAMTS1 IL6 CDH11 0,01987 0.04235 6 PLAU ADD3 INHBA 0,01991 ~~~7~1 004883 7 FGF12 PPL LM02 0.02034 8 NRCAM COL1A2 RElN 002098 9 PPARG MRC1 KOR 002376 10 VEGFC ANGPT2 IL7R 002405 G H 1.8 LEC 1.8 ~ ~ • LEC 11 1.6 •0 BEC 11 1.6 DBEC e 1.4 e 1.4 ** C C n.s *** 0 0 I ~ 1.2 ~ 1.2 * I Q) Q) Cl Cl 1 c C l1l l1l .<: 0.8 .<: 0.8 0 0 -0 -0 :E 0.6 :E 0.6 Q) Q) > 0.4 > 0.4 :; ~ Qi 0.2 Qi 0.2 0:: 0:: 0 0 0 0.05 0.5 5 50 0 0.5 5 50 500 IL-7 (ng/ml) FGF-12 (ng/ml) Figure 3.2.3 Identification of (lymph)angiogenic-mediators using Prediction Relevance Ranking test and ANOVA (A-D) Differential immunofluorescence analysis of psoriatic lesional skin stained for the lymphatic marker LYVE-1 (green; arrows) and for the vascular marker CD31 (red) revealed 4 different patterns: Samples with low blood vessel and lymphatic vessel expansion (A; quadrant IV in panel E), strong blood vessel but low lymphatic vessel expansion (B; quadrant III in panel E), low blood vessel but strong lymphatic vessel expansion (C; quadrant I in panel E) and strong blood vessel and lymphatic vessel expansions (D; quadrant II in panel E). Scale bars: 100 µm. BVA = tissue area covered by blood vessels (in %). LVA = tissue area covered by lymphatic vessels (in %). (F) Comparative analysis of Prediction Relevance Ranking (PRR) and ANOVA ranking, based on mRNA gene expression profiles of 43 psoriasis skin lesions, revealed FLT1, FGF12, IL7R (bold) to be significantly overrepresented in these analyzes as (lymph)angiogenic-mediators. (G, H) Treatment with IL-7 or with FGF-12 for 48 h dose-dependently promoted cellular proliferation of LEC (filled bars) and of BEC (open bars). ***P-value < 0.0005; **P-value < 0.005; *P-value < 0.05).
3.2.3 Discussion
We have established the first complete lineage-specific transcriptome of cultured human LEC and compared it with that of BEC. Analysis of the molecular pathways associated with the transcriptome of each cell type revealed lineage-specific functions. Furthermore, we developed the LD-MDA to quantify expression of vascular lineage-specific genes in various endothelial cell types and we identified two
68 novel mediators of (lymph)angiogenesis that are associated with the extent of vascular expansion in the chronic inflammatory skin disease psoriasis.
Using microarray analyses of three matched pairs of human LEC and BEC, we were able to comprehensively characterize the vascular lineage-specific transcriptome of human endothelial cells, and to identify 342 BEC and 236 LEC signature genes. We reliably detected a number of known BEC-specific (Hirakawa et al, 2003; Hong et al, 2004b; Petrova et al, 2002a) and LEC-specific markers (Prevo et al, 2001; Schacht et al, 2003; Wigle et al, 2002a), as well as a number of previously unknown vascular lineage markers (Hirakawa et al, 2003; Petrova et al, 2002a; Podgrabinska et al, 2002). Several of these genes, such as IL7 and glia maturation factor-gamma, might have important roles in endothelial lineage-specific differentiation and development. In fact, mutations in one of the differentially expressed genes in LEC signatures genes, SOX18, cause recessive and dominant forms of hypotrichosis-lymphedema- telangiectasia syndrome(Irrthum et al, 2003). Although a large number of novel lymphatic-specific and blood vascular-specific gene have been identified, more functional characterization of these genes need to be performed.
In silico analysis of the biological pathways associated with LEC and BEC-specific genes revealed that LEC significantly overexpress genes associated with fatty acid and steroid and cholesterol metabolism. These findings are in agreement with the important role of intestinal lymphatic vessels in the uptake of lipids and with recent results observed in mice that are deficient in the lymphatic-specific transcription factor Prox1. Mice with a targeted disruption of Prox1 in the lymphatic endothelium accumulate fat in lymphatic-rich regions (Harvey et al, 2005). Prox1 might therefore directly control the expression of LEC genes involved in lipid metabolism. It has been previously shown that cholesterol 25-hydroxlyase, one of the genes identified in this study as LEC-specific, was strongly up-regulated in cultured endothelial cells transfected with a Prox1-expressing adenoviral vector (Petrova et al, 2002a).
Previously, endothelial cells derived from large vessels such as HUVEC and from microvessels such as HDMEC were believed to have distinct biological functions (Li et al, 2002; Prabhakarpandian et al, 2001; Unger et al, 2002). Our analysis of these
69 cell types using LD-MDA indicates that differences observed in these cell types might have been caused by varying admixtures of lymphatic and blood vascular endothelial cells in HDMEC preparations, whereas HUVEC are purely of blood vessel origin. So the molecular, transcriptional and functional differences observed between the endothelial cells of small and large vessels should be reevaluated, in light of our findings that LEC represent the majority of cells (at varying percentages, data not shown) in commercially available HDMEC preparations. We advice the routine analysis of the degree of endothelial lineage-specific differentiation that has occurred in each batch of commercial HDMEC, as a quality control step, before these cells are used for research applications. It might also be necessary to evaluate the endothelial cell lines that have been used in previous studies. Using the LD-MDA, we showed that the widely used immortalized HMEC-1 cell line (Ades et al, 1992) has lost several key endothelial-specific characteristics. Results obtained from studies with HMEC-1 should therefore be cautiously interpreted, with regard to their potential relevance to primary endothelial cells. In this regard, the LD-MDA could serve as an easy and reliable tool for quality control analysis of human endothelial cell samples.
Over the recent years, applications of bioinformatics and statistics have generated powerful methods to manage and interpret data, and even to create a statistical model for the prognosis of diseases (Spira et al, 2007; Yu et al, 2008). However, the classical approaches to report only one “best” model failed to characterize the whole model space. High correlations between markers often lead to not only one but also a group of models that perform equally well regarding their predictive power. Hence, the results which report only one of these models are often misleading by ignoring equally good marker combinations. The Prediction Relevance Ranking (PRR) analysis circumvents such problem by counting the number of times each variable shows up in all possible significant models, however, it comes with the drawback of high computation times. Therefore PRR is not suitable for large datasets such as microarrays but it is preferable for biological and clinical datasets which often exhibit a small number of features.
IL7R was identified by PRR and ANOVA as a factor with potential involvement in the mediation of inflammatory angiogenesis and lymphangiogenesis. Previous reports
70 shown that human microvascular endothelial cells express IL7R, the receptor for IL7(Dus et al, 2003). However, our study demonstrates that human microvascular endothelial cells represent a mixture of both lymphatic and blood vascular endothelium. Based on the LD-MDA analyses of a large number of LEC and BEC, we found that IL7R is indeed more strongly expressed by BEC whereas its ligand IL7 is preferentially expressed by LEC. IL7R is also expressed by myeloid lineage cells and can mediate the production of pro-inflammatory cytokines by monocytes/macrophages (Kilroy et al, 2007; Moller et al, 1996), however, the effect of IL7 on endothelial cells has remained unclear. Our functional in vitro studies reveal that IL7 significantly promotes the cellular proliferation of blood vascular and lymphatic endothelium. In this study, we found that FGF12 expression was highly associated with angiogenesis and lymphangiogenesis during chronic inflammation. The specific functions of FGF12 have not yet been determined; however FGF receptors are expressed by vascular endothelium (Suhardja & Hoffman, 2003), and our in vitro functional studies show that FGF12 induces LEC and BEC proliferation. These findings indicate that FGF12 and/or IL7 might represent novel therapeutic targets for the treatment of psoriasis and, possibly, other chronic inflammatory diseases that are characterized by extensive angiogenesis and lymphangiogenesis.
71 3.3 Lymphatic-specific expression of dipeptidyl peptidase IV and its dual role in lymphatic endothelial function
3.3.1 Introduction
The lymphatic vascular system is an open-ended network of endothelial cell-lined vessels that transport extravasated fluid, proteins, metabolites and cells from the interstitial space back to the circulatory system via the thoracic duct (Oliver & Detmar, 2002). Moreover, the lymphatic vascular system also serves as the primary conduit for malignant tumor cell metastasis to regional lymph nodes, and induction of lymphangiogenesis by tumors actively promotes cancer metastasis (Dadras et al, 2005; Hirakawa et al, 2005b; Oliver & Detmar, 2002). There is increasing evidence that lymphatic vessels also actively participate in acute and chronic inflammation. The chronic inflammatory skin disease psoriasis is characterized by pronounced cutaneous lymphatic hyperplasia (Kunstfeld et al, 2004). Kidney transplant rejection is frequently accompanied by lymphangiogenesis (Kerjaschki et al, 2006) and lymphangiogenesis has also been observed in experimental models of chronic airway inflammation (Baluk et al, 2005). However, the molecular mediators of lymphatic vessel activation have remained poorly characterized.
During embryonic development, the transcription factor Prox1 plays a major role in the differentiation and sprouting of lymphatic progenitor cells from the cardinal veins (Hong & Detmar, 2003a). Beginning at embryonic day (E) 9.5 of mouse development, Prox1 is specifically expressed by a subpopulation of endothelial cells that are located on one side of the anterior cardinal vein. These Prox1-positive LECs then bud from the veins to form the primary lymph sacs, which then proliferate and sprout into the periphery to form lymphatic capillaries and vessels (Oliver & Detmar, 2002; Wigle et al, 2002a). Budding and sprouting of LEC from the veins is arrested at ~E11.5-E12.0 in Prox1 null mice (Wigle et al, 2002a). During later stages of development, several genes such as podoplanin (Schacht et al, 2003), neuropilin-2 (Yuan et al, 2002), FOX C2 (Petrova et al, 2004) and angiopoietin-2 (Thurston, 2003) are involved in regulating normal lymphatic vessel patterning and maturation. Although some of these factors are also involved in lymphatic vessel activation under
72 pathological conditions, the mechanisms controlling lymphatic vessel growth and function have remained poorly understood.
Dipeptidyl peptidase IV (DPPIV) is a membrane glycoprotein that cleaves a conserved proline residue in proteotypically resistant components such as collagens, and that regulates the activities of a number of growth factors and neuropeptides (Bauvois, 2004; Busek et al, 2004; Mentlein, 2004). DPPIV is involved in diverse biological processes, including cell differentiation, adhesion, and apoptosis, functions that are important for controlling neoplastic transformation (Boonacker & Van Noorden, 2003; Houghton et al, 1988; Proost et al, 1999; Wesley et al, 1999). In addition, DPPIV mediates binding to collagen (Bauvois, 1988; Loster et al, 1995) and denatured collagen or gelatin (Ghersi et al, 2002). Despite its role in a number of cellular processes, the potential role of DPPIV for the growth and function of the lymphatic vascular system has remained unknown.
Based on transcriptional profiling studies that revealed an increased expression of DPPIV in cultured lymphatic endothelial cells (LEC) as compared to blood vascular endothelial cells (BEC), we aimed to characterize the vascular expression and function of DPPIV. We found– for the first time – that DPPIV expression is specifically expressed by lymphatic vessels but not by blood vessels in skin, as well as in a number of other organs including the small intestine, esophagus, ovary, breast and prostate glands. Studies in primary human LEC revealed that DPPIV is enzymatically active in these cells, but also promotes adhesion to fibronectin and collagen type I, as well as LEC migration and tube formation. These findings identify DPPIV as a novel lymphatic endothelium-specific marker, and they indicate that DPPIV plays a major role in mediating lymphatic endothelial functions.
3.3.2 Results
3.3.2.1 Enhanced expression of DPPIV/CD26 by LEC as compared to BEC To identify genes that are specifically expressed or up-regulated by LEC, as compared to blood vascular endothelial cells (BEC), we isolated and purified both LEC and BEC from human neonatal foreskins of three independent donors. The three LEC and
73 BEC cell lines were then subjected to transcriptional profiling by microarray analysis using Applied Biosystems Human Genome Survey 2.0 (Shin et al., manuscript submitted). These studies revealed that DPPIV/CD26 is expressed at higher levels by LEC than by BEC (23.1-fold average increase; n=3). The difference in DPPIV gene expression was confirmed by quantitative TaqMan real-time RT-PCR in three matched pairs of LEC and BEC that were obtained from the same donor each, with an up to 12-fold increase of DPPIV mRNA levels in LEC (Fig. 3.3.1A). Western blot analyses of cell lysates confirmed that the enhanced mRNA expression levels correlated with enhanced protein expression of DPPIV in LEC (Fig. 3.3.1B).
A 16 • LEC 0 BEC
Line1 Line2 Line3 B BEC LEC BEC LEC
DPPIV ACTB
Figure 3.3.1 Enhanced expression of DPPIV/CD26 by LEC as compared to BEC. (A) Quantitative real-time RT-PCR confirmed that three independently established lines of primary LEC (filled bars) expressed high levels of DPPIV as compared to primary BEC (open bars). (B) Western blot analyses of cell lysates confirmed that LEC expressed much higher levels of DPPIV protein, as compared to BEC (left pane). Western blot analyses for β-Actin were performed for equal loading (right pane).
3.3.2.2 Lymphatic vessels in normal skin specifically express DPPIV To investigate whether DPPIV is also expressed by lymphatic vessels in situ, we next performed double immunofluorescence analyses of normal human skin for DPPIV and for the lymphatic markers LYVE-1, podoplanin and Prox1. LYVE-1-positive (Fig. 3.3.2B), podoplanin-positive (Fig. 3.3.2E) and Prox1-positive (Fig. 3.3.2N) lymphatic vessels also expressed DPPIV ((Fig. 3.3.2A-F and M-O). Immunofluorescent staining for the panendothelial marker CD31 revealed a complete
74 overlap of DPPIV staining with the weakly stained CD31-lymphatic vessels (Fig. 3.3.2G-I), whereas strongly stained CD31-positive blood vessels did not express DPPIV. In agreement with these findings, staining for the blood vascular-specific marker CD34 and for DPPIV was mutually exclusive (Fig. 3.3.2J-L). Taken together, these findings confirm that DPPIV is specifically expressed by lymphatic vessels and not by blood vessels in human skin.
Figure 3.3.2 Specific detection of DPPIV expression by lymphatic endothelium in human skin. Double immunofluorescence analyses of normal human skin for DPPIV (green) and for the lymphatic specific markers (B) LYVE-1, (E) D2-40/podoplanin and (N) Prox1 revealed co-localization (C, F, O). Immunofluorescent staining for the panendothelial marker CD31 revealed a complete overlap of DPPIV staining with the weakly stained CD31-lymphatic vessels (G-I), whereas strongly stained CD31-positive blood vessels did not express DPPIV. Similarly, stainings for DPPIV and for the blood vascular-specific marker CD34 was mutually exclusive (J-L). Scale bars: 100 µm.
75
3.3.2.3 DPPIV is expressed by lymphatic vessels in several human organs We next investigated whether DPPIV might also serve as a specific marker for lymphatic vessels in other human tissues, in addition to the skin. To this end, we analyzed human tissue microarrays containing a number of sections of normal human organs. We found that lymphatic vessels in the small intestine, esophagus, ovary, breast, peripheral nerve tissue and prostate glands expressed DPPIV (Fig. 3.3.3A-O). It is of interest that several glands, including the prostate (Fig. 3.3.3P, Q, R), salivary glands, and adrenal glands (data not shown) showed high expressions of DPPIV by glandular epithelium. Moreover, liver hepatocytes, proximal tubules of the kidney and bile ducts of the liver were also positive for DPPIV (data not shown). In all human tissues examined, DPPIV-positive lymphatic endothelium also expressed the lymphatic-specific marker podoplanin, whereas blood vessels were DPPIV-negative.
76
Figure 3.3.3 DPPIV is expressed by lymphatic vessels in several human organs. We found DPPIV- positive lymphatic vessels in small intestine (A), esophagus (D), cervix (G), breast (J), peripheral nerve (M) and prostate gland (P). Notably, high expressions of DPPIV were found in several glands such as in small intestine (C) and prostate (Q). Scale bars: 100 µm.
77 3.3.2.4 Diprotin A inhibits the enzymatic activity of DPPIV but does not induce LEC proliferation and migration To further characterize the potential functional roles of DPPIV in LEC, we next investigated whether DPPIV produced by LEC is enzymatically active. Using a standard DPPIV activity assay for the cleavage of aminoluciferin, we found that the enzymatic activity of DPPIV was significantly higher in LEC than in BEC (p<0.001), and that the activity increased with increasing cell numbers (Fig. 3.3.4A). The specificity of the enzymatic activity was confirmed by treatment of LEC with the DPPIV-specific inhibitor diprotin A, which resulted in a significant, dose-dependent repression of DPPIV cleavage activity (Fig. 3.3.4B). However, treatment with diprotin A (ranging from 0.01 nM to 10 nM) did not affect LEC proliferation and migration as compared to untreated controls (Fig. 3.3.4C and D).
A B 200 200 *** S- • LEC S- • LEC -.J -.J ~ OBEC ~ 160 OBEC 160 .~ .~ C C :J :J C C 120 Q) 120 Q) ü Ü ifl ifl Q) Q) ~ .~ .~ E 80 E 80 2 2 Q) Q) > > ~ 40 ~ 40 Q) Q) oe oe
20 200 2000 20000 0.1 1 10 100 Cells per cm 2 Diprotin A (nM)
C LEG proliferation D LEG trans-weil migration 2.5 _ 70 ~ :::J -.J " l:S 60 e 2.0 C .~ 0 c 50 ü :J ~ Q) 1.5 ü 40 Ol c c Q) ü .c ifl 30 '"ü 1.0 i" 10 0 :J :Si 20 '"Q) ~ 0.5 > ~ ~ 10 Q) Q) oe 0 oe 0.01 0.1 10 10% FBS 0.01 0.1 10 10% FBS Diprotin A (nM) Diprotin A (nM)
Figure 3.3.4 Diprotin A inhibits the enzymatic activity of DPPIV but does not induce LEC proliferation and migration. (A) The enzymatic activity was significantly higher in LEC (filled circle) than in BEC (open circle) and that the activity increased with increasing cell numbers. (B) Treatment of LEC with the DPPIV-specific inhibitor diprotin A, significantly and dose dependently repressed DPPIV cleavage activity (from 0.1 nM to 1 nM); treatment of BEC with diprotin A slightly repressed DPPIV cleavage activity (from 0.1 nM to 1 nM). (C, D) Furthermore, treatment with diprotin A (ranging from 0.01 nM to 10 nM) did not affect LEC proliferation and LEC migration, whereas 10% FBS significantly induced both LEC proliferation and migration. ***P-value < 0.0001; **P-value < 0.001; *P-value < 0.01.
78
3.3.2.5 SiRNA-mediated knockdown of DPPIV inhibits LEC adhesion, migration and tube-formation In addition - and independently of - its enzymatic activity, DPPIV has also been described to mediate binding to extracellular matrix molecules including fibronectin and collagen (Bauvois, 2004; Loster et al, 1995). We next investigated whether DPPIV inhibition might regulate LEC functions that might play a role in lymphangiogenesis, including cell adhesion, migration and tube formation. Since the inhibition of DPPIV's enzymatic activity by treatment with diprotin A did not affect LEC proliferation and migration, we next aimed to inhibit DPPIV expression by siRNA-mediated knockdown. Using DPPIV-specific siRNA and Amaxa nucleofection, we achieved a > 82% knockdown of DPPIV mRNA expression, as determined by quantitative real-time RT-PCR (Fig. 3.3.5A). DPPIV siRNA knockdown inhibited the adhesion of LEC to both fibronectin and to collagen type I, as compared to control LEC (P<0.005; Fig. 3.3.5B). Trans-well migration assays revealed that LEC transfected with DPPIV siRNA migrated significantly less efficiently towards a FBS gradient when compared to control LEC (P<0.005; Fig. 3.3.5C). Moreover, DPPIV knockdown also inhibited LEC migration in a monolayer scratch wounding assay (P < 0.0005; Fig. 3.3.5D). Knockdown of DPPIV in LEC also inhibited the formation of tube-like structures after overlay of confluent cultures with a collagen type I gel (Fig. 3.3.5G). In contrast, knockdown of DPPIV did not affect LEC proliferation (data not shown).
79 A100% 200 8 120 C
Passage 0 Passage 1 Fibronectin Collagen
E 200000 D G E 8000 :::l. 2: ci- 160000 -5 6000 ~ Cl Cll 120000 c:: ~ ~ Q.'l 4000 Cll 80000 .0 "0 c:: .2 :J 40000 cu 2000 ~ ;§ PBS 3% FBS Control siRNA DPP4siRNA
D ContraI siRNA • DPP4 siRNA
Figure 3.3.5 Knockdown of DPPIV inhibits LEC adhesion, migration and tube formation in vitro. (A) Greater than 82% knockdown of DPPIV mRNA expression was achieved even after passaging the transfected cells once. (B) DPPIV siRNA knockdown inhibited the adhesion of LEC (filled bars) to both fibronectin and to collagen type I, as compared to control LEC (open bars). (C) Trans-well migration assays revealed that LEC transfected with DPPIV siRNA (filled bars) migrated significantly less efficiently towards a FBS gradient when compared to control LEC (open bars). (D) DPPIV knockdown also inhibited LEC migration in a monolayer scratch wounding assay. Cell culture images reveal repressed wound closure in LEC transfected with DPPIV siRNA (F) when compared to control LEC (E). (G) Knockdown of DPPIV in LEC (I) also inhibited the formation of tube-like structures after overlay of confluent cultures with a collagen type I gel when compared to control LEC (H). Scale bars: 100 µm. ***P-value < 0.0005; **P-value < 0.005; *P-value < 0.05.
3.3.3 Discussion
In a search for novel pathways involved in lymphatic vessel growth and function, we have used transcriptional profiling of cultured human dermal BEC and LEC to identify enhanced expression of DPPIV in lymphatic endothelium in vitro. These results were confirmed by quantitative real-time RT-PCR and by Western blot analyses. We also found that DPPIV promotes LEC adhesion, migration and tube formation.
DPPIV has been implicated in several pathological conditions such as rheumatoid arthritis, Grave's disease and tumor progression (Blazquez et al, 1992; Eguchi et al, 1989; Gerli et al, 1996; Hafler et al, 1985; Wesley et al, 1999). Furthermore, recent
80 reports indicated that DPPIV might also play a role in endothelial cells (Chen et al, 2003; Zukowska et al, 2003). In this study, we found– for the first time – that DPPIV is specifically expressed by lymphatic vessels but not by blood vessels in the skin and in a number of additional organs, including the small intestine, esophagus, ovary, breast, and prostate glands. However, DPPIV was not detected on lymphatic vessels in the lung, kidney, uterus, liver and stomach (data not shown).
DPPIV has several functions, including serine peptidase activity, binding to the extracellular matrix, and complexing adenosine deaminase (Bauvois, 2004; De Meester et al, 1999). Each of these distinct functions, presumably mediated by distinct domains, might contribute to its role in lymphatic function. Our results indicate that DPPIV, expressed by LEC, efficiently cleaved the DPPIV substrate Gly-Pro- aminoluciferin, demonstrated that DPPIV expressed in LEC is enzymatically active and functional. DPPIV has the ability to cleave other bioactive peptides such as CXCL12, RANTES, MDC and I-TAC (De Meester et al, 1999; Proost et al, 2000; Proost et al, 2001). Therefore, DPPIV expressed by lymphatic vessels may contribute to the activation or deactivation of chemokines which control trafficking of monocytes, lymphocytes and dendritic cells into lymph nodes via lymphatic vessels. Whereas this enzymatic activity of DPPIV was efficiently inhibited by diprotin A, LEC proliferation and migration were not affected. However, we found that siRNA knockdown of DPPIV significantly inhibited LEC adhesion to fibronectin and collagen type I. These results indicate a dual function of DPPIV in lymphatic endothelium: Whereas the peptidase activity modulates the activity of proinflammatory chemokines and other mediators, DPPIV also mediates the interaction of lymphatic vessels with the extracellular matrix, an essential feature for the efficient drainage function of lymphatic vessels and the interstitial transport of macromolecules (Castenholz, 1998; Oliver & Detmar, 2002; Swartz, 2001). Moreover, siRNA-mediated DPPIV knockdown also inhibited LEC migration and tube formation which are essential for developmental and pathological lymphangiogenesis. These results are in agreement with previous studies which indicated that migration of other cell types was mediated by the adhesive properties of DPPIV (Ghersi et al, 2006; Kertesz et al, 2000). Therefore, specifically targeting the adhesive domain of DPPIV might provide a novel strategy for inhibiting pathological lymphangiogenesis. Future studies are needed to investigate whether DPPIV might
81 also play a role in the mediation of tumor-induced lymphangiogenesis and lymphatic metastasis.
82 3.4 Transcriptional profiling of VEGF-A and VEGF-C target genes in lymphatic endothelium reveals endocan as a novel mediator of lymphangiogenesis
3.4.1 Introduction
The lymphatic vascular system has an important role in the maintenance of tissue fluid homeostasis, in the afferent phase of the immune response, and in acute and chronic inflammation (Alitalo et al, 2005; Cueni & Detmar, 2006a; Kunstfeld et al, 2004). Recent studies have revealed that lymphatic vessels also play an active role in the metastatic spread of malignant tumor cells to regional lymph nodes (Mandriota et al, 2001; Skobe et al, 2001; Stacker et al, 2001). In particular, tumors can induce lymphangiogenesis via release of the lymphangiogenic growth factors VEGF-C or VEGF-D, leading to enhanced rates of metastasis to the draining sentinel lymph nodes and beyond (Mandriota et al, 2001; Skobe et al, 2001; Stacker et al, 2001). Indeed, studies have revealed that tumor-induced lymphangiogenesis is the most significant prognostic indicator to predict the occurrence of regional lymph node metastasis in malignant melanomas of the skin (Dadras et al, 2005). More recently, it has been found that tumors can also induce lymphangiogenesis within their draining lymph nodes, even before they metastasize (Hirakawa et al, 2007; Hirakawa et al, 2005b) and that induction of lymph node lymphangiogenesis promotes the further metastatic cancer spread to distant lymph nodes and to organs (Hirakawa et al, 2007). Thus, tumor-induced lymphatic growth and activation represents a novel potential target for treating or preventing advanced cancer.
Within the last few years, several mediators of lymphangiogenesis have been identified. Hepatocyte growth factor (HGF, also known as scatter factor) was recently found to induce proliferation, migration and tube formation of lymphatic endothelial cells (LEC) and to promote lymphangiogenesis in vivo (Kajiya et al, 2005). Additionally, FGF-2 promotes both lymphatic vessel growth in the mouse cornea (Chang et al, 2004; Kubo et al, 2002), and also promotes proliferation and migration of LEC by binding to its receptor FGFR-3 which is upregulated by the transcription factor Prox1 in lymphatic endothelium (Shin et al, 2006). Other growth factors with
83 effects on the lymphatic vasculature include platelet-derived growth factor-BB, insulin-like growth factor-1 and -2 (Bjorndahl et al, 2005; Cao et al, 2004), angiopoietin-1 (Gerber et al, 1999) and adrenomedullin (Fritz-Six et al, 2008). Despite the growing number of novel potential lymphangiogenic factors, there is strong evidence that growth factors of the vascular endothelial growth factor (VEGF) family, acting via VEGF receptor-3 (VEGFR-3) and VEGFR-2 on lymphatic endothelium, represent the most important lymphangiogenic stimuli in the majority of human and experimental cancers.
VEGF-C promotes lymphangiogenesis by activating VEGFR-2 and VEGFR-3 on LEC (Makinen et al, 2001). VEGF-C-deficient mice fail to develop a functional lymphatic system (Karkkainen et al, 2004), and transgenic expression of a soluble VEGFR-3 results in pronounced lymphedema (Makinen et al, 2001). Recently, VEGF-A has also been implicated as a strong lymphangiogenic mediator. Indeed, adenoviral delivery of murine VEGF-A164 to the skin of mice strongly promoted lymphatic vessel growth, and transgenic mice overexpressing murine VEGF-A164 specifically in the skin show enhanced lymphangiogenesis during wound healing and inflammation (Hirakawa et al, 2005b; Hong et al, 2004b; Kunstfeld et al, 2004; Nagy et al, 2002). Importantly, when VEGF-A transgenic mice were subjected to a standard chemically-induced multistep skin carcinogenesis regimen, there was enhanced proliferation of VEGFR-2-expressing tumor-associated lymphatic vessels, leading to an increased incidence of lymph node metastasis (Hirakawa et al, 2005b). The relative importance of direct VEGF-A induced signaling via VEGFR-2 versus the potential induction of a paracrine stimulatory loop via upregulation of VEGF-C expression by LEC has remained unclear. Moreover, in contrast to the detailed investigation of the effects of VEGF-A on the blood vasculature (Carmeliet, 2003), the downstream targets of VEGF-A (as well as of VEGF-C) in the lymphatic vasculature have remained unknown.
In this study, we aimed to comprehensively identify downstream molecular targets induced by VEGF-A or VEGF-C in lymphatic endothelium. To this end, we treated human dermal microvascular lymphatic endothelial cells (LEC) with VEGF-A or VEGF-C for up to 24 hours, followed by a time-series transcriptional profiling using gene microarray technology. In these studies we identified a number of genes - many
84 of them not previously known to be involved in lymphangiogenesis - that clustered either as early response genes, transiently induced genes or progressively induced genes. Endocan, also known as endothelial specific molecule-1 (ESM-1) was one of the genes that were most potently induced by both VEGF-A and VEGF-C. Whereas ESM-1 induction by VEGF-A was mainly dependent on activation of VEGFR-2, VEGF-C-mediated induction depended on the activity of both VEGFR-2 and VEGFR-3. We found that incubation of LEC with ESM-1 enhances the stimulating effects of both VEGF-A and VEGF-C on LEC proliferation and migration, whereas incubation with ESM-1 alone had no effect. Importantly, VEGF-A (or VEGF-C)- induced induction of LEC proliferation and migration was significantly inhibited by siRNA-mediated silencing of ESM-1 in vitro and in vivo. Together, these studies reveal endocan/ESM-1 as a novel mediator of lymphangiogenesis and as a potential target for the inhibition of VEGF-A- or VEGF-C-induced pathological lymphatic vessel growth and activation.
3.4.2 Results
3.4.2.1 Microarray analysis reveals novel mediators of VEGF-A and VEGF-C- induced effects on lymphatic endothelial cells Both VEGF-A and VEGF-C have been shown to promote lymphangiogenesis in vivo and to enhance lymphatic endothelial cell (LEC) proliferation and migration in vitro (Hirakawa et al, 2007; Hirakawa et al, 2005b; Hong et al, 2004b; Kunstfeld et al, 2004). To identify genes involved in lymphangiogenesis mediated by these factors, we incubated human dermal microvascular LEC with either VEGF-A or VEGF-C for 0, 1, 4, 8 and 24 hours in triplicates, followed by gene microarray analyses using the chemiluminescence-based Applied Biosystems Human Genome Microarrays platform. We investigated the differentially expressed genes by applying multivariate Empirical Bayes statistics, which ranks genes on the basis of their sequential expression over time and the reproducibility at each time point (Friedman et al, 2000; Liang & Kelemen, 2007; Pan, 2002; Zhang & Zhang, 2007). We next performed Short Time Series Expression Miner (STEM) analysis (Ernst & Bar-Joseph, 2006) to determine which significantly modulated genes cluster together based on their
85 temporal regulation pattern (Fig. 3.4.1). For the VEGF-A treated LEC, we identified 71 genes clustering into the early-response genes group (ER; peak at time point 1h), 49 genes into the transiently induced genes group (TI; peak between 4h and 8h), 79 genes into the progressively induced genes group (PI; progressive increase of expression over time) and 52 into the downregulated genes group (DR; progressive decrease over time). For the VEGF-C treated LEC, 74 genes clustered into ER, 41 into TI, 35 into PI and 38 into DR (Fig. 3.4.1). The early response gene cluster revealed the most overlapping genes induced by both VEGF-A and VEGF-C (n=26) as compared to the temporal clusters, and included known early response genes such as EGR-1, EGR-2 and EGR-3 (Table 3.4-1). As previously described for blood vascular endothelium (Hesser et al, 2004), DSCR1 was one of the most strongly induced VEGF-A early response genes.
Early response Transiently induced Progressively induced Downregulated of\ 1/ ~
14 8 24h 14 8 24h 14 8 24h 14 8 24h VEGF-A: 71 genes 49 genes 79 genes 52 genes VEGF-C: 74 genes 41 genes 35 genes 38 genes Common: 26 genes 2 genes 8 genes 7 genes
Figure 3.4.1 Microarray time course analysis of LEC treated with VEGF-A or VEGF-C reveals four major temporally regulated gene clusters. Transcription profiling and Short Time Series Expression Miner (STEM) analysis of LEC stimulated with VEGF-A or VEGF-C for 1h, 4h, 8h or 24h revealed 71/74 genes specifically up-regulated at 1h (early response genes), 49/41 genes upregulated transiently (transiently induced), 79/35 genes upregulated progressively over time (progressively induced) and 52/38 genes down-regulated over time (downregulated).
Among the progressively induced genes, we identified several genes that have been previously reported to be involved in the mediation of lymphangiogenesis, including VEGF-C and angiopoietin-2 (Thurston, 2003; Veikkola et al, 2001). We additionally found upregulation of asp-like, microcephaly associated (ASPM), TTK protein kinase (TTK) and kinesin family member 14 (KIF14) (Table 3.4-1). Fatty acid binding protein 3 (FABP3), SHC SH2-domain binding domain 1 (SHCBP1) and cell division
86 cycle 2 (CDC2) were highly upregulated by both VEGF-A and VEGF-C in LEC. The list of significantly modulated genes is provided in Appendix Table 2 and 3.
Table 3.4-1 Top 10 genes induced by VEGF-A, VEGF-C or both
a) Early response genes b) Transiently induced genes VEGF-A & VEGF-A & VEGF-A VEGF-C VEGF-C VEGF-A VEGF-C VEGF-C NR4A2 EGR3 DUSP5 PLAUR ANKRD20B MYO1B DSCR1 EGR2 LOC387763 LRP8 SLC4A7 LY6H NR4A3 NR4A1 TncRNA PLAT FOXI1 TRIB1 F3 KLF10 GAL ETV1 AXUD1 FOSB PFKFB3 ACOT11 CAMTA1 KLF10 FOS LOC441655 NPAS2 CAMK2B RRAD EGR1 TSC22D2 SHC4 KLF5 NUAK2 PTGS2 DKFZP434F0318 UHRF1 IL18R1 TNFAIP3 ATF3 PFKFB3 FLT1 GUCY1A3 MAP3K8 STC1 PELO SPHK1 CLGN
c) Progressively induced genes d) Dowregulated genes VEGF-A & VEGF-A & VEGF-A VEGF-C VEGF-C VEGF-A VEGF-C VEGF-C ANGPT2 SHCBP1 LOC338579 CA4 PDK4 FGFR1 ASPM CDC2 VEGFC KCTD12 REPS2 CITED1 TOP2A FABP3 C20orf128 CGNL1 CYP1A1 PGLYRP2 KIF14 ZWINT ITGB1BP2 GUCY1A3 C2orf23 SCARF2 ARHGAP11A ESM1 GALNT8 BRUNOL5 LTB TMC8 NUSAP1 CXCR4 NP TGFA TMEM100 UTP14A KIF2C ST8SIA4 LDLRAD1 MAN1C1 SPATA12 LAMA2 DGKD CDC45L MPHOSPH6 LFNG WASF2 DAF CTAGE4 SLC2A12 KCNQ1 TTK CORO6 IGF1 INE1
We next applied the PANTHER annotation and classification software to identify biological pathways with time-specific regulation by VEGF-A. Among the 14 molecular functions significantly overrepresented after 1h of VEGF-A treatment, were transcription factors and signaling molecules (p<0.0005; Table 3.4-2a). Within the transiently and progressively induced gene clusters, genes encoding cytokine receptors and growth factors were significantly overrepresented. The most significantly overrepresented molecular functions after 24 hours included cytoskeletal and microtubule binding proteins (p<0.0005). According to their biological process annotations, genes induced after 1h were significantly involved in mRNA transcription, cell proliferation and cell cycle control (Table 3.4-2b). Among the
87 overrepresented pathways after 24h of VEGF-A treatment were protein modification and phosphorylation. The results for VEGF-C treated LEC are provided in Appendix Table 4.
Table 3.4-2 Pathway classification analysis of VEGF-A induced genes in LEC
Molecular function 1h 4h 8h 24h Cytokine receptor ++ ++ ++ ++ Phosphorylase ++ ++ ++ ++ Growth factor + +++ + + Kinase modulator + + + + Kinase inhibitor + + ++ - Phosphatase ++ + +++ - Basic helix-loop-helix transcription factor + + + - Signaling molecule +++ +++ + - Carbohydrate phosphatase ++ +++ +++ - Kinase + + - +++ Select regulatory molecule + - + + Protein phosphatase + - ++ - Transcription factor +++ - - - Metalloprotease + - - - Defense/immunity protein - + + - RNA-binding protein - + + - Replication origin binding protein - - + ++ Transaminase - - ++ + Transferase - - + - Microtubule binding motor protein - - - +++ Microtubule family cytoskeletal protein - - - +++ Cytoskeletal protein - - - +++ Protein kinase - - - ++ Actin binding motor protein - - - + DNA topoisomerase - - - + DNA strand-pairing protein - - - + DNA helicase - - - +
88
Biological process 1h 4h 8h 24h Cell proliferation and differentiation +++ + + + Developmental processes +++ + + + Nucleoside, nucleotide and nucleic acid metabolism +++ + + + Intracellular signaling cascade ++ ++ + + MAPKKK cascade +++ ++ +++ - Immunity and defense ++ ++ ++ - Receptor protein tyrosine kinase signaling pathway + + ++ - JNK cascade + + + - Signal transduction +++ +++ + - Protein phosphorylation +++ + - +++ Ligand-mediated signaling +++ +++ - - Cell communication +++ ++ - - Cell cycle + - + +++ Cell cycle control +++ - + +++ Protein modification + - - ++ Inhibition of apoptosis ++ - - + Cell surface receptor mediated signal transduction ++ - - - mRNA transcription +++ - - - Cell adhesion-mediated signaling + - - - Neurogenesis + - - - Complement-mediated immunity - + + + Angiogenesis - + + + Amino acid biosynthesis - - + ++ Mitosis - - - +++ Chromosome segregation - - - +++ Cytokinesis - - - +++ DNA replication - - - ++
+++ p-value < 0.0005; ++ p-value < 0.005; + p-value < 0.05; - not significant
3.4.2.2 ESM-1 expression is potently induced in LEC by VEGF-A and VEGF-C Among the progressively increasing gene cluster, endocan (also known as endothelial specific molecule-1; ESM-1), was one of the most potently upregulated genes in LEC after VEGF-A and VEGF-C treatment. Quantitative TaqMan real-time RT-PCR analyses revealed a more than 10-fold induction of ESM-1 mRNA expression at 24h of VEGF-A treatment, and a more than 4-fold induction after treatment with VEGF-C (Fig. 3.4.2A), thus confirming the microarray results. Western blot analyses confirmed that ESM1 protein expression was also strongly increased in both LEC lysates and supernatants at 24h and 48h of VEGF-A treatment, compared to control LEC (Fig. 3.4.2B,C).
Although VEGF receptor-2 (VEGFR-2) is the only presently known receptor for VEGF-A on LEC, some of the observed VEGF-A effects might have been mediated
89 by an indirect pathway involving upregulation of VEGF-C which then might have led to VEGFR-3 activation. Moreover, the mature form of human VEGF-C used for the experiments can bind to both VEGFR-2 and VEGFR-3. Thus, we next investigated the relative contribution of VEGFR-2 and VEGFR-3 towards the induction of ESM-1 by VEGF-A and VEGF-C, using blocking antibodies specific for VEGFR-2 or VEGFR-3. Treatment of LEC with VEGF-A strongly induced the expression of ESM- 1 in the presence of control IgG, whereas ESM-1 induction was completely inhibited by a VEGFR-2 blocking antibody (p<0.0001; Fig. 3.4.2D). Incubation with a VEGFR-3 blocking antibody partially reduced the induction of ESM-1 by VEGF-A (p<0.001), and a combination of both blocking antibodies inhibited the VEGF-A- mediated ESM-1 induction (p<0.0001; Fig. 3.4.2D). VEGF-C also induced the expression of ESM1 in the presence of control IgG, though less potently than VEGF- A (Fig. 3.4.2E). Incubation with either an anti-VEGFR-2 antibody or an anti- VEGFR-3 antibody only partially blocked the ESM-1 induction by VEGF-C (p<0.001 and p<0.01, respectively; Fig. 3.4.2E). Combined blockade of VEGFR-2 and VEGFR-3 completely prevented VEGF-C-mediated induction of ESM-1 expression (p<0.0001).
90 A 16 B Lysate • VEGF-A (qPCR) 25 kDa- 14 -er VEGF-A (microarray) 'C' .c • VEGF-C (qPCR) 0 12 D VEGF-C (microarray) 11 e 15 kDa c 10 0 -2- l3-actin Q) 8 Cl c Supernatant .c'" 6 1J" 0 LL 4 50 kDa-
2 35 kDa-
0 4 8 24 VEGF-Ä o 24 48 hr C Time (hours) D ~ 9 ,------, f. 6,------, " 8 ".2' ~ 5 e 7 e "E 6 "E 4 0 o ~ u 5 ';3 •Cl Cl 0 4 .r:• iii u 3 "5 2 :l!"" 2 :l!"" 1 :E :E Ul 0 Ul 0 W IgG a-R2 a-R3 a-R2 IgG a-R2 a-R3 a-R2 w IgG a-R2 a-R3 a-R2 IgG a-R2 a-R3 a-R2 +u-R3 +u-R3 +u-R3 +u-R3 ---::-::--:------::-:-::--- PBS VEGF-A PBS VEGF-C
E LYVE-1 F ESM-1 G Merge WT - --- - ' - .
H ~ ~- LYVE-1 I ESM-1 J --. ~ - Merge TG - , , , -- -
Figure 3.4.2 Expression of ESM1 in LEC is induced by VEGF-A and VEGF-C via VEGFR-2 and VEGFR-3. (A) When compared to untreated controls, LEC expressed over 10-fold higher levels of ESM1 mRNA after 24h stimulation with VEGF-A (filled circle), and over 4-fold higher levels with VEGF-C (filled square), confirming the microarray expression results (open circles, open squares) by quantitative RT-PCR. (B) Western blot analysis confirmed that LEC stimulated with VEGF-A for 24h or 48h also expressed much higher levels of ESM1 in the cell lysates and in cell supernatants, as compared to untreated controls. (C) Treatment of LEC with VEGF-A strongly induced the expression of ESM-1 in the presence of control IgG, whereas ESM-1 induction was completely inhibited by a VEGFR-2 blocking antibody. Incubation with a VEGFR-3 blocking antibody partially reduced the induction of ESM-1 by VEGF-A, and a combination of both blocking antibodies inhibited the VEGF-A- mediated ESM-1 induction. (D) VEGF-C also induced the expression of ESM1. Incubation with either an anti-VEGFR-2 antibody or an anti-VEGFR-3 antibody only partially blocked the ESM-1 induction by VEGF-C, which was completely prevented by combined blockade of both receptors. ***P-value < 0.0001; **P-value < 0.001; *P-value < 0.01. The LYVE-1 positive lymphatic vessels in the skin of VEGF-A transgenic mice (H-J) expressed ESM-1 (J, arrows), but not the lymphatic vessels in the skin of wildtype mice (E-G). Scale bars: 100 µm.
91 To investigate whether ESM-1 expression by lymphatic endothelium might also be upregulated by VEGF-A in vivo, we next performed differential immunofluorescence analyses of skin samples obtained from VEGF-A transgenic mice for ESM-1 and the lymphatic-specific hyaluronan receptor LYVE-1. VEGF-A transgenic mice express the murine VEGF-A165 under control of the epidermis-specific keratin 14 promoter, leading to enhanced VEGF-A levels within the skin (Kunstfeld et al, 2004; Xia et al, 2003). The subset of LYVE-1 positive lymphatic vessels in the skin of VEGF-A transgenic mice expressed ESM-1 (Fig. 3.4.2H-J), but not the lymphatic vessels in the skin of wildtype mice (Fig. 3.4.2E-G).
3.4.2.3 ESM-1 promotes LEC proliferation and migration induced by VEGF-A and VEGF-C Because both VEGF-A and VEGF-C promoted ESM-1 expression by LEC, we investigated whether ESM-1 might modulate the effects of both growth factors on lymphatic endothelial cell functions. Incubation of LEC with ESM-1 alone did not affect LEC proliferation at concentrations ranging from 0.1 ng/ml to 1 µg/ml (Fig. 3.4.3A and data not shown). However, addition of ESM1 together with VEGF-A or VEGF-C significantly and dose-dependently increased the stimulatory effects of both growth factors on LEC proliferation (Fig. 3.4.3A).
We next investigated whether silencing of ESM-1 expression by small interfering RNAs (siRNAs) might affect the proliferative effects of VEGF-A and -C on LEC. Transfection of LEC with ESM-1 siRNAs efficiently reduced the ESM-1 protein levels, as compared with control siRNA-transfected LEC (Fig. 3.4.3B). When ESM-1 siRNA-transfected LEC were treated with VEGF-A or VEGF-C, the proliferation- inducing effects of both growth factors were potently suppressed, as compared with control siRNA-transfected LEC (p<0.005; Fig. 3.4.3C). Addition of human recombinant ESM-1 protein to ESM-1 siRNA-transfected LEC partially restored the level of growth stimulation by both VEGF-A and VEGF-C (p<0.005 and p<0.05, respectively; Fig. 3.4.3C).
We next investigated whether ESM-1 might also modulate the effects of VEGF-A or - C on LEC migration. Using a standard monolayer wound assay in vitro, we found that
92 addition of ESM-1 at a concentration of 100 ng/ml slightly promoted the migration- enhancing effect of VEGF-A, as compared with VEGF-A treatment only (Fig. 3.4.3D). LEC transfected with ESM-1 siRNA showed a significantly reduced migratory response to VEGF-A treatment (p<0.005), as compared to control siRNA- transfected LEC. Addition of recombinant ESM-1 protein restored the full stimulatory effect of VEGF-A on LEC migration (Fig. 3.4.3D). Comparable results were seen when LEC were treated with VEGF-C (data not shown).
93 A 3.0 • VEGF-A 2.5 D VEGF-C
11 e 2.0 "E 0 .!:!- 1.5 QJ l/) ro QJ Ü 1.0 .~ "C 0 0.5 LL
0 o 1.0 o 0.01 01 1.0 2.5 ESM-1 (lJg/ml) B c • ContraI siRNA D ESM1 siRNA 3 Supernatant
11 2.5 50 kDa- e "E o 2 .!:!- lll1.5 35 kDa- ro ~ ü .~ si RNA e;;;;tr;;'1 ESM1 "C ~ 0.5 o PBS VEGF-A VEGF-A VEGF-C VEGF-C D + ESM-1 + ESM-1 200 E :::l. 180 0- l/) 160 0 0 0 140 E 120 -e-PBS ro -0- PBS (ESM1 siRNA) QJ 100 • VEGF-A ro -0- VEGF-A (ESM1 siR NA) "C c 80 -.- VEGF-A+ ESM1 :::> 0 -fr VEGF-A + ESM1 (ESM1 siRNA) ~ 60 QJ 40 Clro Gi 20 > Figure 3.4.3 ESM1 promotes LEC proliferation and migration induced by VEGF-A and VEGF- C. (A) Addition of ESM1 together with VEGF-A (20 ng/ml) or VEGF-C (100 ng/ml) significantly and dose-dependently increased the stimulatory effects of both growth factors on LEC proliferation, whereas ESM1 alone had no effect. (B) Transfection of LEC with ESM-1 siRNAs reduced ESM-1 protein levels compared with control siRNA-transfected LEC. (C) The proliferation-inducing effects of VEGF-A and VEGF-C were suppressed in ESM-1 siRNA-transfected LEC but not in control siRNA- transfected LEC. Addition of ESM-1 protein to ESM-1 siRNA-transfected LEC partially restored growth stimulation by VEGF-A and VEGF-C. (D) ESM1 (100 ng/ml) slightly promoted the promigratory effect of VEGF-A in a monolayer wound assay, whereas LEC transfected with ESM-1 siRNA showed a significantly reduced migratory response to VEGF-A compared to control siRNA- transfected LEC. Addition of ESM-1 protein restored the effect of VEGF-A on LEC migration. *P- value<0.05; **P-value<0.005; ***P-value<0.0005. 94 3.4.2.4 ESM-1 promotes lymphatic vessel activation by VEGF-A in vivo Because ESM-1 expression by LEC was strongly upregulated by VEGF-A in vitro, and was also upregulated on lymphatic vessels in the skin of VEGF-A transgenic mice, we next investigated whether ESM-1 might also promote the effects of VEGF- A on lymphatic endothelium in vivo. To this end, we used an established Matrigel implantation assay in FVB wildtype mice. Matrigels containing PBS or VEGF-A together with either ESM-1 siRNA or with control siRNA were subcutaneously injected into mice. After 7 days, tissue samples were obtained and frozen sections were subjected to differential immunofluorescence analyses for the lymphatic marker LYVE-1 and the panendothelial marker CD31. In the skin surrounding Matrigel implants containing VEGF-A and control siRNA, LYVE-1-positive lymphatic vessels were strongly enlarged (Fig. 3.4.4B), as compared to Matrigels containing PBS only (Fig. 3.4.4A). In contrast, lymphatic vessels showed a normal morphology and were not enlarged in the skin surrounding Matrigel implants containing VEGF-A and ESM- 1 siRNA (Fig. 3.4.4C). Computer-assisted morphometric analyses of LYVE-1/CD31 stained sections demonstrated that the density of lymphatic vessels was comparable in all three groups (Fig. 3.4.4D). However, the average size of lymphatic vessels was significantly larger in the skin surrounding Matrigel implants containing VEGF-A and control siRNA, as compared to Matrigels containing PBS alone (P<0.005; Fig. 3.4.4E). Importantly, lymphatic vessels in the skin surrounding implants containing VEGF-A and ESM-1 siRNA showed a comparable size as found surrounding PBS- containing implants and were significantly smaller than surrounding VEGF-A/control siRNA implants (P<0.005; Fig. 3.4.4E). Similarly, the average tissue area covered by lymphatic vessels was significantly increased surrounding VEGF-A/control siRNA implants, as compared to PBS-containing (P<0.0005) or VEGF-A/ESM-1 siRNA containing implants (P<0.05; Fig. 3.4.4F). 95 Contral VEGF-Alcontral-siRNA VEGF-AlESM 1-siRNA 2 2 D Lymphatic vessels/mm E Lymphatic vessel size (IJm ) F Lymphatic vessel are (%) 25,------, 1000,------, 1.6 20 800 15 600 10 400 200 Contral VEGF-A VEGF·A Contral VEGF-A VEGF-A Control VEGF-A VEGF-A Ctrl-siRNA ESM1-siRNA Ctrl-siRNA ESM1·siRNA Ctrl·siRNA ESM1-siRNA Figure 3.4.4 Targeting ESM1 by siRNA inhibits VEGF-A effects on lymphatic vessels in vivo. Compared to Matrigels containing PBS (A), LYVE-1-positive lymphatic vessels were enlarged surrounding Matrigels containing VEGF-A and control siRNA (B). Lymphatic vessels were not enlarged surrounding Matrigels containing VEGF-A and ESM-1 siRNA (C). Scale bars: 100 µm. Computer-assisted morphometric analyses of LYVE-1/CD31 stained sections showed comparable density of lymphatic vessels in all groups (D). The average size of lymphatic vessels was significantly larger surrounding Matrigels containing VEGF-A/control siRNA than Matrigels containing PBS (E) or VEGF-A/ESM-1 siRNA. The average tissue area covered by lymphatic vessels was significantly increased surrounding VEGF-A/control siRNA implants, as compared to PBS-containing or VEGF- A/ESM-1 siRNA containing implants (F). *P<0.05; **P<0.005; ***P<0.0005. 3.4.3 Discussion In order to comprehensively identify downstream molecular targets induced by VEGF-A or VEGF-C in lymphatic endothelium, we have treated dermal lymphatic endothelial cells with VEGF-A or VEGF-C for up to 24 hours, followed by a time- series transcriptional profiling using gene microarray technology. To our knowledge, this is the first study into the comprehensive downstream mediators of VEGF-A in lymphatic endothelium. Microarray time course studies provide the ability to monitor the temporal behavior of biological processes of interest through the sequential measurement of the expression of tens of thousands of genes (Martinez et al, 2007; Murphy, 2002; Qian et al, 2003; Storey et al, 2005). However, given the large number of genes evaluated in microarray time course experiments, and the usually rather small number of replicates, the variances are usually poorly estimated. Thus, for the analysis of gene 96 expression of triplicate samples at 5 different time points, we have applied the multivariate Empirical Bayes (EB) approach to inference (Friedman et al, 2000; Liang & Kelemen, 2007). This is a model-based strategy for introducing moderation into the analysis and has been reported to generate the least number of false positives and false negatives (Huber et al, 2002). Using short time series expression miner (STEM) analysis (Ernst & Bar-Joseph, 2006), we identified distinct temporal clusters of genes modulated by VEGF-A and VEGF-C. Within the early response cluster - with a peak of expression after 1h - we predominantly found transcription factors and signaling molecules. Several of these, such as the early growth response genes EGR1, EGR2 and EGFR3 have been previously described as early VEGF-A target genes in blood vascular endothelium (Liu et al, 2003a; Schoenfeld et al, 2004; Yang et al, 2002). Our finding of Down syndrome critical region protein 1 (DSCR1), an inhibitor of NFAT activity, as one of the most potently induced early response targets in LEC is in agreement with previous results in human umbilical vein endothelial cells (Hesser et al, 2004). It is of interest that the overlap among VEGF-A and VEGF-C targets was highest in the early response cluster, as compared to the transiently induced and the progressively induced gene clusters. Among the cluster of genes with progressively increasing expression after VEGF-A and VEGF-C treatment were - angioipoietin-2 which has been previously shown to be indispensable for the normal development of the lymphatic vasculature in mice (Gale et al, 2002) - as well as VEGF-C, indicating the possible induction of an autocrine growth pathway via VEGFR-3 activation (see below). Importantly, we found endocan (also known as endothelial specific molecule-1; ESM- 1) to be one of the most potently induced genes by both VEGF-A and VEGF-C in cultured LEC - both at the mRNA and the protein level. ESM-1 expression was also detectable on lymphatic vessels in the skin of transgenic mice with chronically elevated levels of VEGF-A but not in the skin of wildtype mice, indicating that ESM- 1 is also a target of VEGF-A in vivo. ESM-1 is a dermatan sulphate proteoglycan secreted by endothelial cells that has been suggested to play a role in the regulation of cell adhesion in inflammatory disorders and in tumor progression (Bechard et al, 2001b). Whereas the precise function of ESM-1 is still unclear, it has been proposed to inhibit the interaction between intercellular adhesion molecule-1 (ICAM-1) and the integrin LFA-1 on lymphocytes and monocytes (Bechard et al, 2001b). Recently, 97 increased ESM-1 mRNA expression levels were reported to represent one of the most significant molecular signatures of a poor prognosis in several types of cancer including lung cancer (Grigoriu et al, 2006). Moreover, overexpression of ESM-1 in human embryonic kidney 293 cells, promoted tumor growth in a xenotransplant model in mice (Scherpereel et al, 2003). Our current study reveals that ESM-1 promoted the mitogenic and promigratory activity of both VEGF-A and VEGF-C on cultured LEC, whereas addition of ESM-1 alone did not affect LEC functions in vitro, similar to blood vascular endothelium (Rennel et al, 2007). Moreover, siRNA-mediating ESM-1 silencing inhibited the activation of LEC by VEGF-A and VEGF-C in vitro and by VEGF-A in vivo. Together, the results indicate that VEGF-A/VEGF-C-mediated induction of ESM-1 represents an autocrine, positive feed-back loop to further promote the stimulatory effects of both growth factors on lymphatic endothelium. ESM-1 has been previously shown to bind to HGF and to increase the HGF mediated proliferation of human embryonic kidney cells in a similar way as heparin, and the single dermatan sulfate chain of the molecule - covalently attached to serine 137 - appears to be required for this effect since the nonglycanated form of ESM-1 did not promote the effects of HGF (Bechard et al, 2001a). Thus, it will be of interest to investigate whether ESM-1 might also bind to VEGF-A and VEGF-C to enhance their interaction with VEGFR-2 and/or VEGFR-3 on LEC. The relative contribution of direct activation of VEGFR-2 versus possible indirect effects on VEGFR-3 - via induction of its ligand VEGF-C - towards the lymphangiogenic effects of VEGF-A have remained unclear (Hirakawa et al, 2005a). In our study, we found that inhibition of VEGFR-2 with a blocking antibody completely abrogated the VEGF-A-mediated induction of ESM-1 in LEC, clearly indicating that VEGFR-2 is essential for mediating VEGF-A effects. However, specific blockade of VEGFR-3 resulted in a partial inhibition of the ESM-1 induction by VEGF-A. Together with the observed induction of VEGF-C expression after VEGF-A treatment, these findings indicate that VEGF-A might mediate its lymphangiogenic effects indeed in part via activation of an autocrine loop in LEC which leads to VEGFR-3 activation by VEGF-C. It remains at present unclear whether or not VEGF-A might also exert effects on the formation of heterodimers of 98 VEGFR-2 and VEGFR-3 which might then affect receptor tyrosine phosphorylation (Dixelius et al, 2003). However, one has to keep in mind that the effects of VEGF-A on LEC gene expression were in general stronger than those of VEGF-C, and that the VEGF-A-mediated induction of ESM-1 was only partially blocked by the anti- VEGFR-3 blocking antibody. Our study also demonstrates that both VEGFR-2 and VEGFR-3 are required for the full activity of VEGF-C on LEC gene expression, since inhibition of each receptor alone only partially inhibited the induction of ESM-1 by VEGF-C whereas combined blockade completely abrogated the VEGF-C effect. These findings are in agreement with results demonstrating that the mature form of the VEGF-C protein, which was used for this study, efficiently binds to and activates both receptors (Alitalo et al, 2005). Overall, our studies reveal endocan/ESM-1 as a novel mediator of lymphangiogenesis and as a potential target for the inhibition of VEGF-A- or VEGF-C-induced pathological lymphatic vessel growth and activation. 99 4 CONCLUSIONS AND OUTLOOK 100 4.1 Conclusions In this work, we set out to investigate the molecular mechanisms of lymphatic vascular function during endothelial lineage-specific differentiation and lymphangiogenesis. We demonstrate that the lymphatic-specific transcription factor Prox1 directly binds to putative Prox1 response elements of the FGFR-3 promoter and upregulates the expression of FGFR-3. To validate this finding, we introduced two amino acid substitution mutations into the DNA binding sites of Prospero, the Drosophila homolog of Prox1 (Ryter et al, 2002) and revealed that the mutated Prox1 completely lost its transcriptional activity. Additionally, we confirmed the Prox1 binding to the putative Prox1 binding elements by performing gel electrophoresis mobility shift assays (EMSA) and transcriptional activation/luciferase reporter assays. Immunohistochemical analyses further revealed that many of the Prox1-positive differentiating endothelial cells were positively stained for FGFR-3 in E11.5 mouse embryos. In order to show the functional role of FGFR-3 in LEC, we stimulated the cells with FGF1 and FGF2 and revealed significantly enhanced migration and proliferation of LEC – independently of the VEGF-C/VEGFR-3 pathway in vitro. In order to identify other Prox1-modulated genes or LEC-specific genes, we set out to perform a gene expression profiling of human dermal BEC and LEC using oligonucleotide microarrays. This analysis revealed a novel set of 236 lymphatic signature genes and 342 blood vascular signature genes. Based on these endothelial lineage-specific transcripts, we established a novel Low-Density Microvascular Differentiation Array (LD-MDA) that revealed that commercially available HDMEC (human dermal microvascular endothelial cells) are a mixed cell population of LEC and BEC, and contained 84.1% LEC and 15.9% BEC. Using the LD-MDA together with biostatistical analysis, the investigation of the gene expression profiles of 43 psoriatic skin lesions revealed FGF12 and IL7 as novel (lymph)angiogenic-mediators in chronic inflammation. Furthermore, stimulation of LEC and BEC with FGF12 and IL7 significantly enhanced cell proliferation in vitro. In additional experiments - based on the microarray studies where we identified DPPIV as one of the 236 identified lymphatic signature genes - we found that the 101 active form of dipeptidyl peptidase IV (DPPIV) is more strongly expressed in lymphatic endothelium as compared to blood vascular endothelium in different tissues including skin. Knockdown of DPPIV in LEC significantly repressed cell migration, tube-formation and adhesion to extracellular matrix components. Together, these findings suggest that DPPIV is an essential mediator of lymphangiogenesis. In addition to DPPIV, several other factors such as VEGF-A and VEGF-C have been reported to regulate lymphangiogenesis. Understanding the molecular mechanisms by which VEGF-A and VEGF-C exert their effects on LEC may reveal novel targets for the prevention of tumor-associated lymphangiogenesis and cancer metastasis. To this end, we stimulated cultured human LEC with VEGF-A or VEGF-C for different time periods and then performed gene microarray analyses. We identified a number of genes that were induced by VEGF-A and/or VEGF-C, either transiently (early response genes) or progressively. In particular, we found that endothelial specific molecule-1 (ESM1), also known as endocan, was significantly upregulated by VEGF- A and VEGF-C. In vitro assays revealed that endocan promotes LEC proliferation and migration in concert with VEGF-A and VEGF-C. Furthermore, siRNA-mediated endocan knockdown reduced VEGF-A/-C induced LEC proliferation and migration in vitro and also inhibited VEGF-A-induced lymphatic vessel enlargement in vivo. 102 4.2 Outlook The transcriptional profiling studies have provided exciting new information about novel genes with potential importance for lymphatic vessel growth and/or function. However, only a small fraction of these genes has been characterized thus far. Therefore, further investigations and in vitro/in vivo validations will be needed to evaluate the biological role - and possible importance as therapeutic target - of these genes. In this thesis, we discovered that Prox1 directly binds to the putative Prox1 response elements in the FGFR-3 promoter and upregulates the expression of FGFR-3 during lymphatic reprogramming. However, several other genes have been reported to be involved in the regulation of endothelial lineage-specific differentiation (Hong et al, 2002; Petrova et al, 2002a). Therefore, we plan to systematically locate putative Prox1 response elements in the promoter regions of genes involved in lymphatic reprogramming. A further characterization of these genes will provide more insight into the molecular mechanisms controlling lymphangiogenesis during development. In this study, the establishment of the LD-MDA allowed a sensitive quantification of endothelial lineage-specific differentiation; however, the genes selected for the LD- MDA and the in vivo application of the LD-MDA platform need to be further optimized. We are currently optimizing the platform using ‘core’ signature genes which show a more comparable expression within a group (e.g. LEC or BEC) but more differential expression between the groups. Furthermore, we try to obtain additional samples of psoriatic skin lesions - in addition to the 43 samples studied in this thesis - to enable an improved statistical analysis and tests for biological significances. The increased sample size could also suggest a possibility to apply the LD-MDA as a prognostic tool to quantify the extent of (lymph)angiogenesis in patients suffering from psoriasis or other angiogenesis-associated diseases. Additionally, we are currently investigating other possible functions of DPPIV in lymphatic endothelium. The hypothesis that DPPIV cleaves SDF-1α to regulate the chemoattraction of CXCR4-positive cancer cells such as the human breast carcinoma 103 cell line MDA-MB-231 toward lymphatic endothelial cells, can be tested using transwell migration assays. LEC coated-membranes can be treated or not with diprotin A to regulate the cleavage of SDF-1α, thus creating a gradient of (in)activated SDF-1α to control the motility of CXCR4-positive MDA cancer cells towards LEC. A previous report revealed that in vivo neutralization of CXCL12/CXCR4 interactions significantly impaired metastasis of breast cancer cells to regional lymph nodes and lung (Muller et al, 2001). Therefore, injecting MDA cells orthotopically into the mammary fat pad of mice and controlling lymph node metastasis by intravenously delivering diprotin A might shed a new light into the function of DPPIV in vivo. In this thesis, we reported that expression of endocan is induced by VEGF-A and VEGF-C in LEC and we further demonstrated that knockdown of endocan inhibited LEC proliferation and migration in vitro and lymphatic vessel enlargement in vivo. Recent reports have revealed that increased levels of endocan are observed in the blood of cancer patients (Grigoriu et al, 2006; Scherpereel et al, 2003). Therefore, we are currently planning to investigate the regulation of lymph node metastasis by targeting endocan in experimental tumor models. Overall, we have identified and characterized several genes that control lymphatic function during differentiation and lymphangiogenesis after performing transcriptional profiling. However, the recent discovery of a multi-layer gene regulation suggests that our biological system of interest is more complex than we have imagined (Chen & Rajewsky, 2006; Kedde et al, 2007). Therefore, systematically identifying microRNAs, single nucleotide polymorphisms (SNPs), epigenetic regulators and RNA-binding proteins controlling the molecular mechanisms of lymphatic vascular function would likely give rise to novel therapeutic targets to prevent lymphatic-associated pathologies. 104 5 MATERIALS AND METHODS 105 5.1 In Vitro 5.1.1 Cell culture 5.1.1.1 Isolation of human dermal BEC and LEC Neonatal human foreskins were obtained after routine circumcision. After enzymatic digestion, the epidermis was removed and dermal cells were mechanically released (Richard et al, 1998). CD34-positive BEC were isolated by immunomagnetic purification with an anti-human CD34 antibody (BC Pharmigen, San Diego, CA) conjugated to immunomagnetic beads (Dynal, Lake Success, NY). Thereafter, the remaining CD34-negative cells were incubated with an immunomagnetic beads- conjugated anti-human CD31 antibody (Dynal) to isolate LECs. LECs were seeded onto fibronectin-coated (10 µg/ml; BD Biosciences, Bedford, MA) and were propagated in endothelial cell basal medium (Cambrex Corp.; East Rutherford, NJ ) containing 20% fetal bovine serum, antibiotics, 2 mmol/L L-glutamine (Invitrogen Corp.; Carlsbad, CA), 10 µg/mL hydrocortisone acetate and 2.5 × 10-2 mg/ml N-6,2’- O-dibutyryl adenosine-3,5’-cyclic monophosphate (Sigma-Aldrich; St. Louis, MO). Confluent primary BEC cultures were further purified by immunomagnetic E-selectin selection after 6 hours of stimulation with recombinant human tumor necrosis factor- α as described (Richard et al, 1998). The lineage-specific differentiation was confirmed by real-time RT-PCR for the lymphatic vascular markers Prox1, LYVE-1 and podoplanin, and for the blood vascular endothelial markers VEGFR-1 and VEGF-C, as well as by immunostains for CD31, LYVE-1 and Prox1. 5.1.1.2 Cells Human dermal microvascular endothelial cells (HDMEC) and Human umbilical vein endothelial cell (HUVEC) were obtained from Cambrex (Verviers, Belgium) and PromoCell (Heidelberg, Germany). Commercial human LEC were also obtained from Cambrex. Human IMR91 dermal fibroblasts were obtained from the National Institute of Aging, USA. The immortalized human epidermal keratinocyte line HaCaT was a kind gift of Dr. Norbert Fusenig, German Cancer Research Center, Heidelberg, 106 Germany. The immortalized human microvascular endothelial cell line HMEC-1 (Ades et al, 1992) was obtained from the Center for Disease Control, Atlanta, GA, USA. Stably transfected rat myoblasts expressing human FGFR3 IIIb or FGFR3 IIIc were kind gifts from Dr. Daniel Podolsky, Massachusetts General Hospital (Kanai et al, 1997). HEK293 cells were purchased from American Type Culture Collection (ATCC) (Manassas, VA). Primary endothelial cells used in all experiments were in their early passages (no greater than passage number 10). 5.2 Target validations 5.2.1 Electrophoretic mobility shift assay 5.2.1.1 GST-Prox1-DNA complex Purification of the GST-Prox1 protein was performed as described (Belecky-Adams et al, 1997; Cui et al, 2004). The GST-Prox1 vector, a kind gift from Dr. M. Duncan (Cui et al, 2004), expresses the C-terminal half of Prox1 (the homeodomain and prospero domains) fused to the GST protein. Rosetta bacterial cells (Novagen, San Diego, CA) were transformed with the GST-Prox1 vector or a control GST vector (pGEX-KG). Bacterial cell extracts were prepared using the BugBuster solution (Novagen). The GST and GST-Prox1 proteins were isolated by Glutathione Sepharose 4B beads (Ahmersham Bioscience, Piscataway, NJ). Five micrograms of purified proteins were incubated in 10 mM Tris (pH 7.5), 10 mM MgCl2, 5 mM EDTA (pH 7.5), 10 mM DTT, 2% NP-40, 10% glycerol, 20% sucrose, 5 µg BSA, and 0.2 µg poly(dI:dC) (poly-deoxy-inosinic-deoxy-cytidylic-acid) for 30 minutes at room temperature, together with 0.05 pmole of 32P-labeled probes (wild-type, ctgggctccCACGCCTCTgggaccgcccg; mutant, ctgggctccACTTAAGCTgggaccgcccg). The protein-DNA complex was separated in a 6% native polyacrylamide gel (30% PA solution, 5X TBE, 10% AP, TEMED) in 0.5X TBE at 200V for 1 hour in an ice- slurry, after a pre-run in 0.5X TBE at 150V at room temperature. For competition 107 assays, 100-fold molar excess of the unlabeled probe was added to the incubation mixtures. 5.2.2 Quantification of RNA 5.2.2.1 Detection and quantification of FGF receptor, DPPIV and ESM1 expressions using qRT-PCR Dual-labeled TaqMan probe-based real-time RT-PCRs were performed to quantify the expression of FGF receptors (Hong et al, 2002). The sequences of forward and reverse primers and dual-labeled probes are as follows: FGFR-1 (CTCCCGAGGCGGAACC, TGAGCTCGATCCTCCTTTTCA, FAM- CCACGCCGAGCGAGGGTCAG-TAMRA), FGFR-3 (GTCATGGAAAGCGTGGTGC, CCAAACTTGTTCTCCACGACG, FAM- TCGGACCGCGGCAACTACACC-TAMRA), and β-actin (TCACCGAGCGCGGCT, TAATGTCACGCACGATTTCCC, JOE- CAGCTTCACCACCACGGCCGAG-TAMRA). In addition, conventional RT-PCR was performed for FGFR-3 using forward and reverse primers (GACGGCACACCCTACGTTAC, GGATGCCTGCATACACACTG) that bind to the 7th and 10th exon of human FGFR- 3, respectively, along with primers for β-actin (TGGGACGACATGGAGAAAAT, GAGGCGTACAGGGATAGCAC). An FGFR-3 cDNA clone (Clone ID, 180447) from Invitrogen (Carlsbad, CA) was used as a probe for Northern blot analysis. RT- PCR analyses were performed at least three times with comparable results. The expressions of DPPIV mRNA, ESM1 mRNA were quantified by real-time RT- PCR using the ABI 7900 HT Fast Real-Time PCR System (Applied Biosystems, Foster City, CA). The probes and primers for DPPIV (Hs00175218_m1) and ESM1 (Hs00199831_m1) were pre-designed by Applied Biosystems (Foster City, CA). Each reaction was normalized with the expression of β-actin as an internal control. 108 5.2.3 Protein 5.2.3.1 Detection of DPPIV and ESM1 proteins using Western blotting For Western blot analysis of DPPIV, LEC were homogenized in lysis buffer. The protein concentrations were determined using NanoOrange® Protein Quantification Kit (Molecular Probes, Eugene, OR). The lysates (100 µg of total protein) were then subjected to SDS-polyacrylamide gel electrophoresis (PAGE), using a NuPAGE™ 10% BT Gel, 1.0 mm, 12 well and NuPAGE™ MES SDS Running Buffer (20x) (both Invitrogen, Carlsbad, CA). The proteins were transferred from SDS gels onto a Trans- Blot® Transfer Medium pure nitrocellulose membrane (BioRad, Hercules, CA) for immunoblot analysis. Blocking was performed with 5% non-fat dry milk in 0.1% Tween®20 (Sigma) in PBS (PBS-T) and then immunoblotted with the anti-DPPIV goat polyclonal antibody (0.2 µg/ml, R&D Systems, Minneapolis, MN). Specific binding was detected by ECL Plus Western Blotting Detection System (GE Healthcare, Buckinghamshire, UK). Equal loading was confirmed with an antibody against β-actin (Sigma, St. Louis, MO). For Western blot analysis of ESM1, serum-starved (0.1% BSA) LEC were stimulated with 20 µg/ml of VEGF-A for 24h and 48h. The cell lysates were homogenized in lysis buffer and the supernatants were collected at each time point. The protein concentration were determined using NanoOrgane® Protein Quantification Kit (Molecular Probes). The lysates (100 µg total protein) and the supernatants (50 µg) were then subjected to the same method as mentioned above. The membrane was then immunoblotted with the human anti-ESM1 goat polyclonal antibody (0.2 µg/ml, R&D Systems, Minneapolis, MN). Specific binding was detected by ECL Plus Western Blotting Detection System (GE Healthcare, Buckinghamshire, UK). Equal loading was confirmed with an antibody against β-actin (Sigma, St. Louis, MO). 5.2.4 In situ expression validations 5.2.4.1 Tissue samples 109 Samples of human skin were from routine circumcision of neonatal foreskin (Massachusetts General Hospital, Boston, MA). Samples of lesional psoriatic skin were obtained by 8 mm punch biopsy from 43 patients with chronic plaque-type psoriasis (mean age 51.95 years; range from 20 to 76 years) after informed consent was obtained. Approval for this study was obtained from the Human Ethics Committee of the Medical University of Vienna and of the University of Kiel. Samples were cut in half, and one half was stored in RNAlater (Ambion, Austin, Texas) for further RNA isolation. The other half was embedded in optimal cutting temperature (OCT) compound (Sakura Finetek; Torrance, CA, USA) and frozen on dry ice for immunofluorescence analyses. 5.2.4.2 Immunofluorescence staining of FGFR-3 in human skin and mouse embryo Immunofluorescence stainings were performed on frozen sections of 4%- paraformaldehyde-fixed neonatal human foreskin sections or on E11.5 mouse embryo sections as previously described (Hong et al, 2002), using antibodies against human FGFR-3 (MAB 7661, R&D Systems Inc., Minneapolis, MN), mouse FGFR-3 (MAB 710, R&D Systems Inc.), or LYVE-1 (Upstate, Charlottesville, VA). Secondary antibodies labeled with AlexaFluor488 or AlexaFluor594 (Molecular Probes, Eugene, OR) were used to detect respective primary antibodies. Nuclei were counter-stained with 20 µg/ml Hoechst bisbenzimide. 5.2.4.3 Immunofluorescence of psoriatic skin Double immunofluorescence analyses of lymphatic vessels and blood vessels were performed on 8-µm cryostat sections as described (Kunstfeld et al, 2004), using a rabbit polyclonal antibody to the lymphatic-specific hyaluronan receptor LYVE-1 (Upstate/Millipore; Billerica, MA) and a mouse monoclonal antibody to the vascular marker CD31 (Dako Cytomation; Glostrup, Denmark), and corresponding secondary antibodies labeled with AlexaFluor488 or AlexaFluora594 (Invitrogen/Molecular 110 Probes; Carlsbad, CA). Nuclei were counterstained with 20 µg/ml of Hoechst bisbenzimide. Sections were examined using an Axioscope 2 mot plus (Carl Zeiss AG; Feldbach, Switzerland) and images were captured with a Zeiss AxioCam MRc. Computer-assisted morphometric vessel analyses of representative LYVE-1 and CD31 double-stained sections, including the determination of the relative tissue area covered by lymphatic vessels ("lymphatic vessel area"; LVA) or by blood vessels ("blood vessel area"; BVA) were performed as described (Kunstfeld et al, 2004). 5.2.4.4 Immunostains of DPPIV Differential immunofluorescence stains using an antibody against DPPIV (1:100, R&D systems) together with antibodies against lymphatic-specific or blood vessel- specific markers were performed on 8-µm cryostat sections as described (Kunstfeld et al, 2004). Stainings were performed using antibodies against the lymphatic-specific hyaluronan receptor LYVE-1 (1:1000; Upstate/Millipore, Billerica, MA), the lymphatic-specific glycoprotein podoplanin (D2-40; 1:200; Signet, Dedham, MA), the lymphatic-specific transcription factor Prox-1 (Covance, Princeton, NJ), the panendothelial marker CD31 (1:100, Dako Cytomation, Glostrup, Denmark) or the blood vessel-specific marker CD34 (1:100, BD Pharmingen, San Diego, CA), and corresponding secondary antibodies labeled with AlexaFluor488 or AlexaFluora594 (Invitrogen/Molecular Probes). Nuclei were counterstained with 20 µg/ml of Hoechst bisbenzimide. Immunohistochemical stains were performed on tissue arrays of normal human tissues (MaxArray human normal tissue microarray slides, Zymed, San Francisco, CA) as described previously (Dadras et al, 2003). Briefly, the primary antibody against DPPIV (1:100) was applied, followed by incubation with conjugated anti-goat immunoglobulin using the 3-amino-9-ethylcabazole peroxidase kit (Vector laboratories, Burlingame, CA). Sections were examined using an Axioscope 2 mot plus (Carl Zeiss AG; Feldbach, Switzerland) and images were captured with a Zeiss AxioCam MRc. For staining of lymphatic vessels, the anti-human podoplanin antibody D2-40 (Schacht et al, 2005) (Signet) was used. 111 5.2.5 Cell culture-based (in vitro) assays 5.2.5.1 Construction of mutant Prox1 and FGFR-3 reporter gene luciferase assays To construct a mutant Prox1, two amino acid substitution mutations (N625A and R627A) were introduced into pcDNA/Prox1 (Hong et al, 2002) by using the QuickChange II site-directed Mutagenesis kit (Stratagene, La Jolla, CA). DNA sequences of the primers used for the mutagenesis reaction are CTCATCAAGTGGTTTAGCgcTTTCgccGAGTTTTACTAC and CTGAATGTAGTAAAACTCggcGAAAgcGCTAAACCACTTG. The resulting product (pcDNA/MutProx1) was sequenced to confirm the base-pair changes. The mouse FGFR-3 promoter-luciferase constructs were kindly provided by Dr. David Ornitz (McEwen & Ornitz, 1998). Each luciferase construct was co-transfected into HEK 293 cells in combination with pcDNA (Invitrogen), pcDNA/Prox1 or pcDNA/MutProx1. Forty-eight hours after transfection, 50 µl of the cell lysates were used to measure the activity of firefly luciferase using the Dual-Glo Luciferase Assay System (Promega, Madison, WI). Another 50 µl of the cell lysates was used to measure the protein concentration by using the Bio-Rad Protein Assay (Bio-Rad, Hercules, CA). Luciferase activity was normalized by the total protein amount. The assays were performed in triplicates in three independent experiments. 5.2.5.2 Binding and internalization of 125I-FGF-2 FGF-2 was labeled with 125I-Na using iodogen (Pierce, Rockford, IL) as a coupling agent according to the manufacturer's instruction. The specific activity of 125I-FGF-2 was 150,000 cpm/ng. FGF-2 binding to high and low affinity sites was investigated as described (Moscatelli, 1987). Cells were seeded at 2.5 x 105/ cm2 and were cultured in complete medium in 3.5-cm diameter dishes for two days. Cells were washed twice with ice cold PBS and were incubated with the indicated concentrations of 125I-FGF-2 in DMEM containing 20 mM Hepes (pH 7.4) and 0.15 % gelatin for 2 h at 4°C. Cells were then washed three times with cold PBS. 125I-FGF-2 was dissociated from its cellular low affinity binding sites by two 20-second washes with ice cold 20 mM Hepes (pH 7.4), 2 M NaCl, and from its high affinity sites by two 20-second washes 112 with ice cold 20 mM NaAc (pH 4.0), 2 M NaCl. Bound 125I-FGF-2 was quantified using a Kontron MR 250 gamma-counter (Saint-Quentin-Yvelines, France). Non- specific binding was determined by incubating LECs in separate dishes with 125I-FGF- 2 and a 100-fold excess of unlabeled ligand. Specific binding was determined by subtracting non-specific binding from total binding. Experiments were done in duplicates and repeated twice with comparable results. Internalization experiments were performed as described (Perollet et al, 1998). Cells in 3.5-cm diameter dishes were incubated with 10 ng/ml of 125I-FGF-2 and shifted to 37°C. After the specified time points (0-24 h), cells were washed 3 times with PBS and twice for 20 seconds with 20 mM Hepes (pH 7.4) containing 2M NaCl and twice for 20 seconds with ice cold 20 mM NaAc (pH 4.0) containing 2M NaCl, to remove cell surface-associated radioactivity. Cells were then extracted with 5 % Triton X-100, 2 % sodium dodecyl sulfate in PBS pH 7.4 and internalized 125I-FGF-2 was quantified by radioactive counting in a Kontron MR 250 gamma-counter. Experiments were done in duplicates and repeated twice. 5.2.5.3 Cell proliferation, migration, apoptosis assays and functional inhibition of FGFR-3 Recombinant human FGF-1 and FGF-2 were purchased from R&D Systems. For proliferation assays, 1,500 LECs were seeded into a fibronectin-coated well of 96 well plates in complete growth medium (Hirakawa et al, 2003). After 24 hours, cells were treated or not with FGFs (10 ng/ml) for 48 hours in low serum medium (2% FBS) containing heparin (1 µg/ml). Cell proliferation was assessed by the MUH fluorescence assay as previously described (Detmar et al, 1990). For migration assays, 24-well FluoroBlok inserts (Falcon, Franklin Lakes, NJ; 8 µm pore size) were coated on the bottom side with 10 µg/ml fibronectin (BD Bioscience) for 1 hour, and then by 100 µg/ml BSA (Sigma) for 1 hour. 750 µl EBM containing 0.2% BSA and heparin (1 µg/ml), supplemented with or without FGFs (10 ng/ml), was added to the bottom chambers. 5x104 LECs in serum-free EBM medium (Clonetics, Watersville, MD) containing 0.2% BSA were added into each well. After 3 hours, cells migrated onto the bottom side of the inserts were stained with Calcein-AM (Molecular Probes) and the fluorescence intensity was measured using the Victor2 Fluorometer (PerkinElmer, 113 Boston, MA). For VEGFR-3 blocking experiments, LECs were pre-incubated with a control IgG or a rat anti-human VEGFR-3 blocking antibody (1 µg/ml) (kindly provided by Dr. Bronek Pytowsky, ImClone Systems Inc., New York, NY) for 10 min. The serum-free EBM media in the bottom chambers contained VEGF-C (100 ng/ml, R&D System) or FGF (10 ng/ml). For apoptosis assays, 4,000 LECs were seeded into a fibronectin-coated well of 96 well plates and cultured for 24 hours. Cells were then incubated for 24 hours in medium containing 0.1% BSA, 20% FBS, 1 µg/ml heparin, with or without FGF-1 or FGF-2 at a concentration of 10 ng/ml. Cytoplasmic histone-associated-DNA-fragments generated by induction of cell death was quantified using the Cell Death Detection ELISA kit (Roche, Indianapolis, IN). Functional inhibition of FGFR-3 was performed by transfecting cultured LECs (passage 2) with pooled small interfering RNAs (siRNA) for FGFR-3 or siRNA for the luciferase gene as a negative control by using Amaxa HMVEC-L Nucleofector Kit (Amaxa Inc., Cologne, Germany). The siRNA sequences are as follow (FGFR-3: CACGACCUGUACAUGAUCAdTdT, UGCACAACGUCACCUUUGAdTdT and UGCACAACCUCGACUACUAdTdT; Luciferase, CUUACGCUGAGUACUUCGAdTdt). Transfected cells were then plated into two 6- cm dishes. One dish was used to collect total RNAs to quantify the steady-state level of FGFR-3 and the other for cell proliferation assays. Proliferation assays were performed 24 hours after transfection as described above. 5.2.5.4 LEC and BEC proliferation assays for FGF-12 and IL7 Recombinant human FGF-12 and IL-7 were purchased from R&D Systems (Minneapolis, MN). BEC and LEC (1.5x103) were seeded into fibronectin-coated 96- well plates and were incubated in complete growth medium (Hirakawa et al, 2003). After 24 hours, cells were incubated in medium containing 1% FBS overnight and quintuplicate wells were then treated or not with FGF-12 (0.5 ng/ml – 500 ng/ml) or IL-7 (0.05 ng/ml – 50 ng/ml) in low serum medium (1% FBS). After 48 h, cells were incubated with 4-methylumbelliferyl heptanoate (MUH; Sigma, St. Louis, MO) as described(Detmar et al, 1990). The intensity of fluorescence, proportional to the number of viable cells, was measured using a SpectraMax Gemini EM microplate 114 reader (Bucher Biotec AG, Basel, Switzerland). Experiments were repeated three times for each treatment. Statistical analyses were performed using the two-tailed unpaired student's t-test. 5.2.5.5 DPPIV enzyme activity assay LEC or BEC were seeded into fibronectin-coated wells of 96-well plates in complete growth medium, at a cell density ranging from 20 to 20,000 cells/cm2. After 24 hours, cells were washed twice with PBS, incubated with the proluminescent DPPIV substrate Gly-Pro-aminoluciferin (DPPIV-Glo protease assay; Promega, Madison, WI), and gently mixed using a plate shaker at 500 rpm for 30 seconds. Plates were incubated for 2 hours in a buffer system optimized for DPPIV and luciferase activities. Luciferase activity was assessed by a LMAXII 384 luminometer (Bucher Biotec AG, Basel, Switzerland). For enzyme inhibition assays, 2,000 LEC were seeded into fibronectin-coated wells of 96-well plates in complete medium. After 20 hours, cells were washed twice with PBS and were treated with diprotin A (International Peptides, Osaka, Japan) for 4 hours. 5.2.5.6 LEC transwell migration, scratch-wound, tube formation and adhesion assays and functional inhibition of DPPIV siRNA-transfection was performed using Basic Nucleofector Kit for primary mammalian endothelial cells (Amaxa Biosystems, Cologne, Germany) according to the manufacturer’s protocol. Predesigned siRNAs against human DPPIV (SI00030212, SI00030219, SI00030226; Qiagen, Hilden, Germany) and control siRNA (Silencer Negative Control #1 siRNA, Ambion, Cambridgeshire, UK) were used for the transfections. For endothelial cell migration assays, control or DPPIV siRNA-transfected LEC were grown to 100% confluency and serum starved overnight. The following day, a cell- free wound zone was created by scraping the monolayer with a sterile pipette tip. The cells were washed with PBS and then the medium was changed to EBM containing either PBS or 3% FBS. The monolayers were incubated in 5% CO2 at 37°C for 48 h. 115 Representative images were taken at 5x magnification directly after wounding and after 48h, using an AxioCam MRm camera attached to an Axiovert 200M microscope (Carl Zeiss AG, Feldbach, Switzerland). Computer-assisted morphometric wound area analyses were performed using the IP-LAB software (Scanalytics, Fairfax, VA). For trans-well migration assays, 24-well FluoroBlock inserts of 8 µm pore size (BD Bioscience, Bedford, MA) were coated on the bottom side with 10 µg/ml fibronectin (BD Biosciences) or with type I collagen (Vitrogen, Palo Alto, CA) for 1 h, followed by incubation with 100 µg/ml bovine serum albumin (BSA; Sigma) to block the remaining protein-binding sites. Cells (1x106 cells/ml; 100 µl) were seeded in serum- free EBM medium (Cambrex Bio Science) containing 0.2% delipidized BSA into the upper chambers, and were incubated for 3 h at 37°C in the presence or absence of 3% FBS. Cells on the underside of inserts were stained with Calcein AM (Molecular Probes), and the fluorescence intensity was measured using a Spectra Max Gemini fluorescence reader (Bucher Biotec AG). Cell adhesion assays were performed by coating 96-well plates with fibronectin (10 µg/ml) or type I collagen (50 µg/ml) for 30 min, followed by blocking with 100 µg/ml BSA. Control or DPPIV siRNA- transfected LEC (105 cells in 200 µl of serum-free EBM) were seeded into each well and were incubated at 37°C for 45 min. Unattached cells were removed by three gentle washes with serum-free EBM containing 0.5% BSA; attached cells were stained with Calcein AM, fixed with 4% paraformaldehyde, and fluorescence was measured using a Spectra Max Gemini EM. Tube formation assays were performed as described (Schacht et al, 2003). Control or DPPIV siRNA transfected LEC were grown on collagen-coated 24 well plates until confluence. Then, 0.5 ml of neutralized isotonic bovine dermal collagen type I (Vitrogen, Palo Alto, CA) with 3% FBS was added to the cells. After incubation at 37°C for 6 h, cells were fixed with 4% paraformaldehyde for 30 min at 4°C. Representative images were captured and the total length of tube-like structures per area was measured using the IP-LAB software (Scanalytics, Fairfax, VA) as described (Schacht et al, 2003). All studies were repeated three times. Statistical analyses were performed using the unpaired Student’s t-test. 116 5.2.5.7 Functional inhibition of ESM-1, receptor blocking experiment and LEC proliferation and migration assays siRNA-transfection was performed using the Basic Nucleofector Kit for primary mammalian endothelial cells (Amaxa Biosystems, Cologne, Germany) according to the manufacturer’s protocol. The following siRNAs against ESM1 were used (sense): 5’-GGUUUGUAAAAGAAGAAUCtt-3’, 5’-GGUGUCAGCCUUCUAAUGGtt-3’ and 5’-GCUGCAUAAGCUGUUAGGUtt-3’, as well as control siRNA (silencer negative control #1 siRNA, Ambion, Cambridgeshire, UK). Antibodies against the extracellular domain of human VEGFR-2 (1121b) (Lu et al, 2000) and of human VEGFR-3 (hF4-3C5) (Persaud et al, 2004), as well as rat negative control IgG were kindly provided by Dr. Bronek Pytowski, Imclone Systems, New York, NY. LEC (2x103) were seeded onto fibronectin-coated 96-well plates. Quintuplicate wells were treated with different concentrations of recombinant human ESM1 (0.01 ng/ml to 2500 ng/ml; R&D Systems) and with 20 ng/ml VEGF-A or 100 ng/ml VEGF-C. In some experiments, cells were also incubated with an anti-human ESM1 polyclonal antibody (10 µg/ml) or with control rat-IgG (10 µg/ml). After 72 hours, cells were incubated with 4-methylumbelliferyl heptanoate (MUH; Sigma-Aldrich). The fluorescence intensity, proportional to the number of viable cells, was measured using a Spectra Max GEMINI EM fluorescence reader (Bucher Biotec AG, Basel, Switzerland). For endothelial cell migration assays, control- or ESM-1 siRNA- transfected LEC were grown to 100% confluency and serum starved overnight. The following day, a cell-free wound zone was created by scraping the monolayer with a sterile pipette tip. The cells were washed with PBS and then the medium was changed to EBM/0.1% BSA containing either PBS or VEGF-A (20 ng/ml) or ESM1 (1 µg/ml) or both. The cells were incubated in 5% CO2 at 37°C for 48h. Representative images were taken at 5x magnification directly after wounding and after 48h, using an AxioCam MRm camera attached to an Axiovert 200M microscope (Carl Zeiss AG). Computer-assisted morphometric wound area analyses were performed using the IP- LAB software (Scanalytics, Fairfax, VA). All experiments were performed three times. Statistical analyses were performed using the unpaired Student’s t-test. 117 5.3 In Vivo 5.3.1 Mouse experiments 5.3.1.1 ESM-1 siRNA Matrigel assay, immunofluorescence stainings and morphometric analyses Lymphangiogenesis was evaluated in vivo by using a matrigel plug assay as described previously (Chae et al, 2004; Kajiya et al, 2005; Zhang et al, 2006). FVB wild-type mice (female, 6-8 weeks old) were anaesthetized and injected subcutaneously into the lower flank skin with 100 µl of Matrigel (BD Biosciences, Bedford, MA) containing either human VEGF-A165 (500 ng/ml) and Silencer Negative control siRNA (10 µg/ml; Ambion, cat. No. 4635) or VEGF-A165 and murine ESM1 siRNA (10 µg/ml; Ambion, cat. no. 16804) (n=5 per group). After 7 days, skin samples were embedded in optimal cutting temperature compound (OCT; Sakura Finetek, Torrance, CA). Immunofluorescence analyses were performed on 8 µm cryostat sections as described (Hong et al, 2004b; Kunstfeld et al, 2004), using a rabbit polyclonal antibody against mouse LYVE-1 (Upstate Biotechnology, Charlottesville, VA) and a monoclonal rat antibody against mouse CD31 (BD Biosciences Pharmingen, San Diego, CA). For the detection of ESM1 in the skin of VEGF-A transgenic mice (female, 8 weeks old) (Kunstfeld et al, 2004; Xia et al, 2003), murine ESM-1 antibody (R&D Systems) was used. Corresponding secondary antibodies were labeled with AlexaFluor488 or AlexaFluor594 (Molecular Probes, Eugene, OR). Nuclei were counterstained with 20 µg/ml Hoechst 33342 (Molecular Probes). Sections were examined by an Axioskop2 microscope (Carl Zeiss AG, Feldbach, Switzerland) and images were captured at 20x magnification with an AxioCam MRm digital camera. Computer-assisted morphometric vessel analyses of representative LYVE-1 and CD31 double-stained sections were performed using the IP-LAB software (Scanalytics, Fairfax, VA) as described (Kunstfeld et al, 2004). Three individual fields per section were examined and the number of vessels per mm2, the average vessel size and the average tissue area covered by vessels were determined. Statistical analysis was performed using the unpaired Student's t-test. 118 5.4 TRANSCRIPTOMICS 5.4.1 Gene expression profiling using oligonucleotide microarrays 5.4.1.1 Gene expression profiling of human LEC and BEC Total cellular RNA was isolated from confluent BEC and LEC cultures after 2−5 passages using the Trizol reagent (Invitrogen). Digoxigenin-UTP labeled cRNA was generated and linearly amplified from 1 µg of RNA for each sample using the Chemiluminescent RT-IVT Labeling Kit v.2.0 (Applied Biosystems; Foster City, CA) according to the manufacturer’s protocol. We obtained the gene expression profiles of three matched pairs (each pair obtained from the same donor) of cultured LEC and BEC by the Applied Biosystems Human Genome Survey Microarray v2.0. Labeling, hybridization, and signal generation and detection were performed according to the manufacturers’ protocols. 5.4.1.2 LEC vs. BEC microarray data analysis Quantile normalization, implemented in the statistical language R (http://www.bioconductor.org/), was applied to the datasets to normalize the distribution of probe set intensities for each array. Present calls were set by a signal- to-noise ratio (S/N) ≥3 and quality-flag ≤ 5,000 determined by the AB1700 microarray software tool. The ratio of gene expression was calculated for each matched pair of LEC and BEC, and was expressed as log2 values. For the identification of the LEC-specific transcriptome, probes with present calls in the LEC samples were selected for further analysis. For the identification of the BEC-specific transcriptome, the same filtering was performed on BEC samples. LEC-specific signature genes were identified based on a log2 ratio (LEC/BEC) ≥ 1, whereas BEC- specific signature genes were selected based on a log2 ratio ≤ − 1 in each of the three matched pairs of LEC and BEC. 119 5.4.1.3 Gene expression profiling for VEGF-A and VEGF-C target genes in LEC Primary LEC were serum starved overnight in EBM supplemented with 0.2% bovine serum albumin. Cells were treated or not for 1h, 4h, 8h or 24h with recombinant human VEGF-A165 (R&D Systems; 20 ng/ml) or mature human VEGF-C (R&D Systems; 500 ng/ml). Total cellular RNA was isolated using the Trizol reagent (Invitrogen) and was extracted with chloroform, precipitated with isopropanol, washed with 70% ethanol, and dissolved in DNase-free/RNase-free distilled water. The concentration of RNA was measured using a NanoDrop ND-1000 spectrophotometer (Witec AG, Littau, Switzerland), and RNA quality was assessed using a 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA). Digoxigenin-UTP labeled cRNA was generated, amplified from 500 ng of total RNA, using the NanoAmp RT-IVT Labeling Kit (Applied Biosystems, Foster City, CA) following the manufacturer's protocol, and was hybridized to Applied Biosystems Human Genome Survey Microarrays V2.0. Chemiluminescence detection, image acquisition and analysis were performed using the Chemiluminescence Detection Kit (Applied Biosystems) and the Applied Biosystems 1700 Chemiluminescent Microarray Analyzer following the manufacturer's protocol. Three biological replicates were generated for each treatment condition (VEGF-A and VEGF-C) and for each time point (0h, 1h, 4h, 8h, 24h). 5.4.1.4 VEGF-A, VEGF-C stimulated time course data analysis Raw data were normalized using Variance Stabilization and Normalization (VSN), a model derived from the variance-versus-mean dependence for microarray intensity data (Huber et al, 2002), available from R/Bioconductor (Gentleman et al, 2004). In a second step, probes which had a signal to noise ratio (S/N ratio) ≥ 3, flag (error) value ≤ 5000 in at least two out of the three replicates for each time point were further subjected to statistical analyses. Differentially expressed genes were identified using the multivariate Empirical Bayes (EB) analysis (R package: time course). The multivariate-EB procedure focuses on moderating the denominator of the multivariate t-statistics, and ranks genes according to the moderated statistic to reduce the number 120 of false positives and false negatives resulting from very small or very large replicate variances or covariances (www.stat.berkeley.edu/tech-reports/667.pdf). In a next step, Time Series Expression Miner (STEM) (Ernst & Bar-Joseph, 2006) was used to identify early response, transiently upregulated, progressively induced and downregulated clusters. Briefly, STEM implements a clustering method that depends on a set of distinct and representative short temporal expression profiles and each probe in the dataset is assigned to a profile with the closest match. The expected number of probes assigned to each profile is estimated by permutation and the statistically significantly overexpressed (p<0.05) profiles are then identified. The preprocessed datasets of three independent experiments were imported into STEM. Experimental profiles with a minimal correlation of 0.7 with the predetermined model profiles were then clustered together. 5.4.1.5 Establishment of a Low-Density Microvascular Differentiation Array Guided by the gene array results, we selected 54 LEC-specific genes and 31 BEC- specific genes, based upon their consistent and strong specific expression in LEC or BEC, as well as on their assignment to important biological pathways. In addition, the five pan-endothelial cell marker genes PECAM-1, vWF, KDR, TEK, CDH5 and the six endogenous control genes ACTB, GAPDH, PGK1, PPIA, RPLP0 and S18 were included in the design of the LD-MDA (Appendix Table 1). The 384-well micro-fluidic cards were produced with pre-designed primer pairs and FAM-labeled TaqMan probes for each of the 96 selected genes in duplicate, with two sample reservoirs per card. The LD-MDA was then used to evaluate the lineage- specific differentiation of a total of 10 independent lines of primary human dermal LEC, of eight independent lines of primary human dermal BEC, of two independent lines of HUVEC cells, of the immortalized human microvascular endothelial cell line HMEC-1, of the immortalized human epidermal keratinocyte line HaCaT, and of primary human dermal fibroblasts. After extraction of total RNA, the mRNA expression levels of the 96 genes were analyzed by quantitative RT-PCR using the 7900HT Real-Time PCR System (Applied Biosystems). The cDNAs were reverse transcribed from 40 ng of total RNA per sample using random primers provided with 121 the High Capacity cDNA Archive Kit (Applied Biosystems) at 25°C for 10 minutes, followed by incubation at 37°C for 120 minutes. PCR products were synthesized using the TaqMan Universal PCR Master Mix (Applied Biosystems) at 95°C for 10 minutes, followed by 40 cycles at 95°C for 15 seconds and 60°C for 60 seconds. In additional experiments, total RNA was obtained from frozen samples of 43 psoriatic lesions, using homogenization with a TissueLyser (Qiagen, Hilden, Germany) and RNA extraction with the Trizol reagent. 5.4.1.6 Endothelial lineage score analysis and identification of core signature genes To quantitatively analyze the expression of vascular signature genes, the cycle values (Ct) of PCR amplification were acquired after 40 cycles using the Applied Biosystems SDS 2.2 software. The threshold was set to 0.2 units of fluorescence intensity. Unamplified samples were assigned a Ct value of 41. Normalization of Ct values was performed as follows: ΔCtgene = Ctgene − Ctβ-actin because out of the six endogenous controls tested, β-actin showed the most consistent expression across different samples. Next, we implemented the ‘sum-clustering’ method to allocate each sample based on the degree of lymphatic or blood vascular endothelial differentiation. To this end, the sum of the ΔCt values of all 54 LEC-specific genes was subtracted from the sum of the ΔCt values of all 31 BEC-specific genes. This value was defined as the endothelial lineage-specific score (ELS). As a measure for the degree of endothelial cell differentiation, the sum of the ΔCt values of the 5 pan-endothelial marker genes was calculated for each sample and defined as vascular lineage score (VLS). As a next step, we aimed to identify a subset of the 85 LD-MDA signature genes that had the most consistent lineage-specific ‘core genes’. Based on the data obtained from ten LEC and eight BEC cultures, the following standard statistics values were calculated for each gene: delta mean Ct value between the LEC group and the BEC group (∆CtL-B), the mean square (MS) value within the LEC and BEC groups (MSwithin), and the MS value between the LEC group and the BEC group (MSBetween). The genes with lower ratios of MSBetween/ MSwithin (F-value) and the lower value of 122 ∆CtL-B were identified as those that were expressed in both LEC and BEC with less specificity. Thus, 19 out of 85 vascular lineage genes that had the lowest F-values and/or ∆CtL-B were removed. The Unweighted Pair Group Method with Arithmetic Mean (UPGMA) (Sokal, 1965) hierarchical clustering was performed using the Spotfire DecisionSite 8.0 software. 5.4.2 Bioinformatics 5.4.2.1 Microarray data mining tools Pathway analyses were performed using the PANTHER (Protein Analysis THrough Evolutionary Relationships) protein classification system (www.pantherdb.org) which classifies proteins into families/sub-families, molecular functions, biological processes and biological pathways. The functional annotation of LEC and BEC signature genes and pathways which were statistically overrepresented in the group of genes upregulated after VEGF-A or VEGF-C treatments were calculated by a random overlapping p-value using binomial tests, with all of the genes represented on the Applied Biosystems Human Genome Survey Microarray serving as the reference list (Cho & Campbell, 2000). 5.4.2.2 Prediction Relevance Ranking (PRR) analysis To investigate the influence of the 90 endothelial-signature genes on lymphatic vessel area and blood vessel area, we used a novel method denoted “Prediction Relevance Ranking” (PRR). It employs multiple linear regression analysis but in contrast to heuristic subset selection methods or regularization techniques like LASSO(Roth, 2004), it enumerates all possible models and investigates the whole model space to produce a ranking of variables based on their predictive power. To prevent overfitting and to estimate the predictive power of each model, 70% of the samples were randomly chosen for training the model and the remaining 30% samples were used for testing; this procedure was repeated 20 times for each model to estimate the distribution of the prediction error. The predictive power was assed by calculating the residual sum of squares (RSS). To characterize the model space for multiple 123 regression with a continuous target variable (e.g. lymphatic vessel area), PRR uses the linear model which contains the intercept as a reference point. Then, the error distribution of each model, which was lower RSS than average, was compared to the error distribution of the reference model by means of paired Student’s t-test. The model with the lowest p-value is the single best model. Furthermore, the set of models with a p-value below a predefined threshold (0.05) was further characterized. In order to measure the prediction relevance of a given variable in relation to all other variables, each variable (gene) in a significant set of models was counted and plotted (Appendix Figure 1). 124 6 BIBLIOGRAPHY 125 Bibliography Abtahian F, Guerriero A, Sebzda E, Lu MM, Zhou R, Mocsai A, Myers EE, Huang B, Jackson DG, Ferrari VA, Tybulewicz V, Lowell CA, Lepore JJ, Koretzky GA, Kahn ML (2003) Regulation of blood and lymphatic vascular separation by signaling proteins SLP-76 and Syk. Science 299(5604): 247-251 Achen MG, Jeltsch M, Kukk E, Makinen T, Vitali A, Wilks AF, Alitalo K, Stacker SA (1998) Vascular endothelial growth factor D (VEGF-D) is a ligand for the tyrosine kinases VEGF receptor 2 (Flk1) and VEGF receptor 3 (Flt4). Proc Natl Acad Sci U S A 95(2): 548-553 Ades EW, Candal FJ, Swerlick RA, George VG, Summers S, Bosse DC, Lawley TJ (1992) HMEC-1: establishment of an immortalized human microvascular endothelial cell line. J Invest Dermatol 99(6): 683-690 Al-Rawi MA, Watkins G, Mansel RE, Jiang WG (2005) The effects of interleukin-7 on the lymphangiogenic properties of human endothelial cells. Int J Oncol 27(3): 721- 730 Alarid ET, Rubin JS, Young P, Chedid M, Ron D, Aaronson SA, Cunha GR (1994) Keratinocyte growth factor functions in epithelial induction during seminal vesicle development. Proc Natl Acad Sci U S A 91(3): 1074-1078. Alitalo K, Tammela T, Petrova TV (2005) Lymphangiogenesis in development and human disease. Nature 438(7070): 946-953 Ambrose CT (2006) Immunology's first priority dispute--an account of the 17th- century Rudbeck-Bartholin feud. Cell Immunol 242(1): 1-8 Angeli V, Ginhoux F, Llodra J, Quemeneur L, Frenette PS, Skobe M, Jessberger R, Merad M, Randolph GJ (2006) B cell-driven lymphangiogenesis in inflamed lymph nodes enhances dendritic cell mobilization. Immunity 24(2): 203-215 Auguste P, Javerzat S, Bikfalvi A (2003a) Regulation of vascular development by fibroblast growth factors. Cell Tissue Res 314(1): 157-166 Auguste P, Javerzat S, Bikfalvi A (2003b) Regulation of vascular development by fibroblast growth factors. Cell Tissue Res Avantaggiato V, Orlandini M, Acampora D, Oliviero S, Simeone A (1998) Embryonic expression pattern of the murine figf gene, a growth factor belonging to platelet-derived growth factor/vascular endothelial growth factor family. Mech Dev 73(2): 221-224 Backhed F, Crawford PA, O'Donnell D, Gordon JI (2007) Postnatal lymphatic partitioning from the blood vasculature in the small intestine requires fasting-induced adipose factor. Proc Natl Acad Sci U S A 104(2): 606-611 126 Baluk P, Tammela T, Ator E, Lyubynska N, Achen MG, Hicklin DJ, Jeltsch M, Petrova TV, Pytowski B, Stacker SA, Yla-Herttuala S, Jackson DG, Alitalo K, McDonald DM (2005) Pathogenesis of persistent lymphatic vessel hyperplasia in chronic airway inflammation. J Clin Invest 115(2): 247-257 Bammler T, Beyer RP, Bhattacharya S, Boorman GA, Boyles A, Bradford BU, Bumgarner RE, Bushel PR, Chaturvedi K, Choi D, Cunningham ML, Deng S, Dressman HK, Fannin RD, Farin FM, Freedman JH, Fry RC, Harper A, Humble MC, Hurban P, Kavanagh TJ, Kaufmann WK, Kerr KF, Jing L, Lapidus JA, Lasarev MR, Li J, Li YJ, Lobenhofer EK, Lu X, Malek RL, Milton S, Nagalla SR, O'Malley J P, Palmer VS, Pattee P, Paules RS, Perou CM, Phillips K, Qin LX, Qiu Y, Quigley SD, Rodland M, Rusyn I, Samson LD, Schwartz DA, Shi Y, Shin JL, Sieber SO, Slifer S, Speer MC, Spencer PS, Sproles DI, Swenberg JA, Suk WA, Sullivan RC, Tian R, Tennant RW, Todd SA, Tucker CJ, Van Houten B, Weis BK, Xuan S, Zarbl H (2005) Standardizing global gene expression analysis between laboratories and across platforms. Nat Methods 2(5): 351-356 Banerji S, Ni J, Wang SX, Clasper S, Su J, Tammi R, Jones M, Jackson DG (1999) LYVE-1, a new homologue of the CD44 glycoprotein, is a lymph-specific receptor for hyaluronan. J Cell Biol 144(4): 789-801 Bates DO, Harper SJ (2002) Regulation of vascular permeability by vascular endothelial growth factors. Vascul Pharmacol 39(4-5): 225-237 Bauvois B (1988) A collagen-binding glycoprotein on the surface of mouse fibroblasts is identified as dipeptidyl peptidase IV. Biochem J 252(3): 723-731 Bauvois B (2004) Transmembrane proteases in cell growth and invasion: new contributors to angiogenesis? Oncogene 23(2): 317-329 Bechard D, Gentina T, Delehedde M, Scherpereel A, Lyon M, Aumercier M, Vazeux R, Richet C, Degand P, Jude B, Janin A, Fernig DG, Tonnel AB, Lassalle P (2001a) Endocan is a novel chondroitin sulfate/dermatan sulfate proteoglycan that promotes hepatocyte growth factor/scatter factor mitogenic activity. J Biol Chem 276(51): 48341-48349 Bechard D, Scherpereel A, Hammad H, Gentina T, Tsicopoulos A, Aumercier M, Pestel J, Dessaint JP, Tonnel AB, Lassalle P (2001b) Human endothelial-cell specific molecule-1 binds directly to the integrin CD11a/CD18 (LFA-1) and blocks binding to intercellular adhesion molecule-1. J Immunol 167(6): 3099-3106 Belecky-Adams T, Tomarev S, Li HS, Ploder L, McInnes RR, Sundin O, Adler R (1997) Pax-6, Prox 1, and Chx10 homeobox gene expression correlates with phenotypic fate of retinal precursor cells. Invest Ophthalmol Vis Sci 38(7): 1293-1303 Bikfalvi A, Dupuy E, Inyang AL, Fayein N, Leseche G, Courtois Y, Tobelem G (1989) Binding, internalization, and degradation of basic fibroblast growth factor in human microvascular endothelial cells. Exp Cell Res 181(1): 75-84 127 Bikfalvi A, Savona C, Perollet C, Javerzat S (1998) New insights in the biology of fibroblast growth factor-2. Angiogenesis 1(2): 155-173 Bjorndahl M, Cao R, Nissen LJ, Clasper S, Johnson LA, Xue Y, Zhou Z, Jackson D, Hansen AJ, Cao Y (2005) Insulin-like growth factors 1 and 2 induce lymphangiogenesis in vivo. Proc Natl Acad Sci U S A 102(43): 15593-15598 Blazquez MV, Madueno JA, Gonzalez R, Jurado R, Bachovchin WW, Pena J, Munoz E (1992) Selective decrease of CD26 expression in T cells from HIV-1-infected individuals. J Immunol 149(9): 3073-3077 Boonacker E, Van Noorden CJ (2003) The multifunctional or moonlighting protein CD26/DPPIV. Eur J Cell Biol 82(2): 53-73 Breitling R (2006) Biological microarray interpretation: the rules of engagement. Biochim Biophys Acta 1759(7): 319-327 Busek P, Malik R, Sedo A (2004) Dipeptidyl peptidase IV activity and/or structure homologues (DASH) and their substrates in cancer. Int J Biochem Cell Biol 36(3): 408-421 Byzova TV, Goldman CK, Jankau J, Chen J, Cabrera G, Achen MG, Stacker SA, Carnevale KA, Siemionow M, Deitcher SR, DiCorleto PE (2002) Adenovirus encoding vascular endothelial growth factor-D induces tissue-specific vascular patterns in vivo. Blood 99(12): 4434-4442 Cabioglu N, Yazici MS, Arun B, Broglio KR, Hortobagyi GN, Price JE, Sahin A (2005) CCR7 and CXCR4 as novel biomarkers predicting axillary lymph node metastasis in T1 breast cancer. Clin Cancer Res 11(16): 5686-5693 Cao R, Bjorndahl MA, Religa P, Clasper S, Garvin S, Galter D, Meister B, Ikomi F, Tritsaris K, Dissing S, Ohhashi T, Jackson DG, Cao Y (2004) PDGF-BB induces intratumoral lymphangiogenesis and promotes lymphatic metastasis. Cancer Cell 6(4): 333-345 Cao R, Brakenhielm E, Li X, Pietras K, Widenfalk J, Ostman A, Eriksson U, Cao Y (2002) Angiogenesis stimulated by PDGF-CC, a novel member in the PDGF family, involves activation of PDGFR-alphaalpha and -alphabeta receptors. FASEB J 16(12): 1575-1583 Cao Y, Linden P, Farnebo J, Cao R, Eriksson A, Kumar V, Qi JH, Claesson-Welsh L, Alitalo K (1998) Vascular endothelial growth factor C induces angiogenesis in vivo. Proc Natl Acad Sci U S A 95(24): 14389-14394 Carmeliet P (2003) Angiogenesis in health and disease. Nat Med 9(6): 653-660 Carmeliet P, Ng YS, Nuyens D, Theilmeier G, Brusselmans K, Cornelissen I, Ehler E, Kakkar VV, Stalmans I, Mattot V, Perriard JC, Dewerchin M, Flameng W, Nagy A, Lupu F, Moons L, Collen D, D'Amore PA, Shima DT (1999) Impaired myocardial 128 angiogenesis and ischemic cardiomyopathy in mice lacking the vascular endothelial growth factor isoforms VEGF164 and VEGF188. Nat Med 5(5): 495-502 Castenholz A (1998) Functional microanatomy of initial lymphatics with special consideration of the extracellular matrix. Lymphology 31(3): 101-118 Cavanagh LL, Von Andrian UH (2002) Travellers in many guises: the origins and destinations of dendritic cells. Immunol Cell Biol 80(5): 448-462 Chae SS, Paik JH, Furneaux H, Hla T (2004) Requirement for sphingosine 1- phosphate receptor-1 in tumor angiogenesis demonstrated by in vivo RNA interference. J Clin Invest 114(8): 1082-1089 Chang LK, Garcia-Cardena G, Farnebo F, Fannon M, Chen EJ, Butterfield C, Moses MA, Mulligan RC, Folkman J, Kaipainen A (2004) Dose-dependent response of FGF- 2 for lymphangiogenesis. Proc Natl Acad Sci U S A 101(32): 11658-11663 Chen JJ (2007) Key aspects of analyzing microarray gene-expression data. Pharmacogenomics 8(5): 473-482 Chen K, Rajewsky N (2006) Natural selection on human microRNA binding sites inferred from SNP data. Nat Genet 38(12): 1452-1456 Chen WT, Kelly T, Ghersi G (2003) DPPIV, seprase, and related serine peptidases in multiple cellular functions. Curr Top Dev Biol 54: 207-232 Cho RJ, Campbell MJ (2000) Transcription, genomes, function. Trends Genet 16(9): 409-415 Cook T, Pichaud F, Sonneville R, Papatsenko D, Desplan C (2003) Distinction between color photoreceptor cell fates is controlled by Prospero in Drosophila. Dev Cell 4(6): 853-864 Couzin J (2006) Genomics. Microarray data reproduced, but some concerns remain. Science 313(5793): 1559 Crawford PA, Gordon JI (2005) Microbial regulation of intestinal radiosensitivity. Proc Natl Acad Sci U S A 102(37): 13254-13259 Cueni LN, Detmar M (2006a) New insights into the molecular control of the lymphatic vascular system and its role in disease. J Invest Dermatol 126(10): 2167- 2177 Cueni LN, Detmar M (2006b) New insights into the molecular control of the lymphatic vascular system and its role in disease. J Invest Dermatol 126: 2167-2177 Cui W, Tomarev SI, Piatigorsky J, Chepelinsky AB, Duncan MK (2004) Mafs, Prox1, and Pax6 can regulate chicken betaB1-crystallin gene expression. J Biol Chem 279(12): 11088-11095 129 Curtis RK, Oresic M, Vidal-Puig A (2005) Pathways to the analysis of microarray data. Trends Biotechnol 23(8): 429-435 Dadras SS, Lange-Asschenfeldt B, Velasco P, Nguyen L, Vora A, Muzikansky A, Jahnke K, Hauschild A, Hirakawa S, Mihm MC, Detmar M (2005) Tumor lymphangiogenesis predicts melanoma metastasis to sentinel lymph nodes. Mod Pathol 18(9): 1232-1242 Dadras SS, Paul T, Bertoncini J, Brown LF, Muzikansky A, Jackson DG, Ellwanger U, Garbe C, Mihm MC, Detmar M (2003) Tumor lymphangiogenesis: a novel prognostic indicator for cutaneous melanoma metastasis and survival. Am J Pathol 162(6): 1951-1960 Dai M, Wang P, Boyd AD, Kostov G, Athey B, Jones EG, Bunney WE, Myers RM, Speed TP, Akil H, Watson SJ, Meng F (2005) Evolving gene/transcript definitions significantly alter the interpretation of GeneChip data. Nucleic Acids Res 33(20): e175 Daynes RA, Spangrude GJ, Roberts LK, Krueger GG (1985) Regulation by the skin of lymphoid cell recirculation and localization properties. J Invest Dermatol 85(1 Suppl): 14s-20s De Meester I, Korom S, Van Damme J, Scharpe S (1999) CD26, let it cut or cut it down. Immunol Today 20(8): 367-375 Dell KR, Williams LT (1992) A novel form of fibroblast growth factor receptor 2. Alternative splicing of the third immunoglobulin-like domain confers ligand binding specificity. J Biol Chem 267(29): 21225-21229 Detmar M, Brown LF, Claffey KP, Yeo KT, Kocher O, Jackman RW, Berse B, Dvorak HF (1994) Overexpression of vascular permeability factor/vascular endothelial growth factor and its receptors in psoriasis. J Exp Med 180(3): 1141-1146 Detmar M, Imcke E, Ruszczak Z, Orfanos CE (1990) Effects of recombinant tumor necrosis factor-alpha on cultured microvascular endothelial cells derived from human dermis. J Invest Dermatol 95(6 Suppl): 219S-222S. Dixelius J, Makinen T, Wirzenius M, Karkkainen MJ, Wernstedt C, Alitalo K, Claesson-Welsh L (2003) Ligand-induced vascular endothelial growth factor receptor-3 (VEGFR-3) heterodimerization with VEGFR-2 in primary lymphatic endothelial cells regulates tyrosine phosphorylation sites. J Biol Chem 278(42): 40973-40979 Dumont DJ, Jussila L, Taipale J, Lymboussaki A, Mustonen T, Pajusola K, Breitman M, Alitalo K (1998) Cardiovascular failure in mouse embryos deficient in VEGF receptor-3. Science 282(5390): 946-949 Dus D, Krawczenko A, Zalecki P, Paprocka M, Wiedlocha A, Goupille C, Kieda C (2003) IL-7 receptor is present on human microvascular endothelial cells. Immunol Lett 86(2): 163-168 130 Dvorak HF, Brown LF, Detmar M, Dvorak AM (1995) Vascular permeability factor/vascular endothelial growth factor, microvascular hyperpermeability, and angiogenesis. Am J Pathol 146(5): 1029-1039 Eguchi K, Ueki Y, Shimomura C, Otsubo T, Nakao H, Migita K, Kawakami A, Matsunaga M, Tezuka H, Ishikawa N, et al. (1989) Increment in the Ta1+ cells in the peripheral blood and thyroid tissue of patients with Graves' disease. J Immunol 142(12): 4233-4240 Eichmann A, Corbel C, Jaffredo T, Breant C, Joukov V, Kumar V, Alitalo K, le Douarin NM (1998) Avian VEGF-C: cloning, embryonic expression pattern and stimulation of the differentiation of VEGFR2-expressing endothelial cell precursors. Development 125(4): 743-752 Enholm B, Karpanen T, Jeltsch M, Kubo H, Stenback F, Prevo R, Jackson DG, Yla- Herttuala S, Alitalo K (2001) Adenoviral expression of vascular endothelial growth factor-C induces lymphangiogenesis in the skin. Circ Res 88(6): 623-629 Ernst J, Bar-Joseph Z (2006) STEM: a tool for the analysis of short time series gene expression data. BMC Bioinformatics 7: 191 Fang J, Dagenais SL, Erickson RP, Arlt MF, Glynn MW, Gorski JL, Seaver LH, Glover TW (2000) Mutations in FOXC2 (MFH-1), a forkhead family transcription factor, are responsible for the hereditary lymphedema-distichiasis syndrome. Am J Hum Genet 67(6): 1382-1388 Ferrara N (2004) Vascular endothelial growth factor: basic science and clinical progress. Endocr Rev 25(4): 581-611 Ferrara N, Gerber HP, LeCouter J (2003) The biology of VEGF and its receptors. Nat Med 9(6): 669-676 Friedman N, Linial M, Nachman I, Pe'er D (2000) Using Bayesian networks to analyze expression data. J Comput Biol 7(3-4): 601-620 Fritz-Six KL, Dunworth WP, Li M, Caron KM (2008) Adrenomedullin signaling is necessary for murine lymphatic vascular development. J Clin Invest 118(1): 40-50 Gale NW, Prevo R, Espinosa J, Ferguson DJ, Dominguez MG, Yancopoulos GD, Thurston G, Jackson DG (2007) Normal lymphatic development and function in mice deficient for the lymphatic hyaluronan receptor LYVE-1. Mol Cell Biol 27(2): 595- 604 Gale NW, Thurston G, Hackett SF, Renard R, Wang Q, McClain J, Martin C, Witte C, Witte MH, Jackson D, Suri C, Campochiaro PA, Wiegand SJ, Yancopoulos GD (2002) Angiopoietin-2 is required for postnatal angiogenesis and lymphatic patterning, and only the latter role is rescued by Angiopoietin-1. Dev Cell 3(3): 411- 423 131 Gasparo A (1627) De lactibus sive lacteis venis. Milan: Mediolani Gentleman RC, Carey VJ, Bates DM, Bolstad B, Dettling M, Dudoit S, Ellis B, Gautier L, Ge Y, Gentry J, Hornik K, Hothorn T, Huber W, Iacus S, Irizarry R, Leisch F, Li C, Maechler M, Rossini AJ, Sawitzki G, Smith C, Smyth G, Tierney L, Yang JY, Zhang J (2004) Bioconductor: open software development for computational biology and bioinformatics. Genome Biol 5(10): R80 Gerber HP, Vu TH, Ryan AM, Kowalski J, Werb Z, Ferrara N (1999) VEGF couples hypertrophic cartilage remodeling, ossification and angiogenesis during endochondral bone formation. Nat Med 5(6): 623-628 Gerli R, Muscat C, Bertotto A, Bistoni O, Agea E, Tognellini R, Fiorucci G, Cesarotti M, Bombardieri S (1996) CD26 surface molecule involvement in T cell activation and lymphokine synthesis in rheumatoid and other inflammatory synovitis. Clin Immunol Immunopathol 80(1): 31-37 Ghersi G, Dong H, Goldstein LA, Yeh Y, Hakkinen L, Larjava HS, Chen WT (2002) Regulation of fibroblast migration on collagenous matrix by a cell surface peptidase complex. J Biol Chem 277(32): 29231-29241 Ghersi G, Zhao Q, Salamone M, Yeh Y, Zucker S, Chen WT (2006) The protease complex consisting of dipeptidyl peptidase IV and seprase plays a role in the migration and invasion of human endothelial cells in collagenous matrices. Cancer Res 66(9): 4652-4661 Gottlieb AB, Chamian F, Masud S, Cardinale I, Abello MV, Lowes MA, Chen F, Magliocco M, Krueger JG (2005) TNF inhibition rapidly down-regulates multiple proinflammatory pathways in psoriasis plaques. J Immunol 175(4): 2721-2729 Grigoriu BD, Depontieu F, Scherpereel A, Gourcerol D, Devos P, Ouatas T, Lafitte JJ, Copin MC, Tonnel AB, Lassalle P (2006) Endocan expression and relationship with survival in human non-small cell lung cancer. Clin Cancer Res 12(15): 4575- 4582 Groger M, Loewe R, Holnthoner W, Embacher R, Pillinger M, Herron GS, Wolff K, Petzelbauer P (2004) IL-3 induces expression of lymphatic markers Prox-1 and podoplanin in human endothelial cells. J Immunol 173(12): 7161-7169 Groth C, Lardelli M (2002) The structure and function of vertebrate fibroblast growth factor receptor 1. Int J Dev Biol 46(4): 393-400 Gunn MD, Tangemann K, Tam C, Cyster JG, Rosen SD, Williams LT (1998) A chemokine expressed in lymphoid high endothelial venules promotes the adhesion and chemotaxis of naive T lymphocytes. Proc Natl Acad Sci U S A 95(1): 258-263 Gunther K, Leier J, Henning G, Dimmler A, Weissbach R, Hohenberger W, Forster R (2005) Prediction of lymph node metastasis in colorectal carcinoma by expressionof chemokine receptor CCR7. Int J Cancer 116(5): 726-733 132 Hafler DA, Fox DA, Manning ME, Schlossman SF, Reinherz EL, Weiner HL (1985) In vivo activated T lymphocytes in the peripheral blood and cerebrospinal fluid of patients with multiple sclerosis. N Engl J Med 312(22): 1405-1411 Hanneken A (2001) Structural characterization of the circulating soluble FGF receptors reveals multiple isoforms generated by secretion and ectodomain shedding. FEBS Lett 489(2-3): 176-181 Harvey NL, Srinivasan RS, Dillard ME, Johnson NC, Witte MH, Boyd K, Sleeman MW, Oliver G (2005) Lymphatic vascular defects promoted by Prox1 haploinsufficiency cause adult-onset obesity. Nat Genet 37(10): 1072-1081 Hassan B, Li L, Bremer KA, Chang W, Pinsonneault J, Vaessin H (1997) Prospero is a panneural transcription factor that modulates homeodomain protein activity. Proc Natl Acad Sci U S A 94(20): 10991-10996. He Y, Rajantie I, Ilmonen M, Makinen T, Karkkainen MJ, Haiko P, Salven P, Alitalo K (2004) Preexisting lymphatic endothelium but not endothelial progenitor cells are essential for tumor lymphangiogenesis and lymphatic metastasis. Cancer Res 64(11): 3737-3740 He Y, Rajantie I, Pajusola K, Jeltsch M, Holopainen T, Yla-Herttuala S, Harding T, Jooss K, Takahashi T, Alitalo K (2005) Vascular endothelial cell growth factor receptor 3-mediated activation of lymphatic endothelium is crucial for tumor cell entry and spread via lymphatic vessels. Cancer Res 65(11): 4739-4746 Heresi GA, Wang J, Taichman R, Chirinos JA, Regalado JJ, Lichtstein DM, Rosenblatt JD (2005) Expression of the chemokine receptor CCR7 in prostate cancer presenting with generalized lymphadenopathy: report of a case, review of the literature, and analysis of chemokine receptor expression. Urol Oncol 23(4): 261-267 Hesser BA, Liang XH, Camenisch G, Yang S, Lewin DA, Scheller R, Ferrara N, Gerber HP (2004) Down syndrome critical region protein 1 (DSCR1), a novel VEGF target gene that regulates expression of inflammatory markers on activated endothelial cells. Blood 104(1): 149-158 Heydtmann M, Hardie D, Shields PL, Faint J, Buckley CD, Campbell JJ, Salmon M, Adams DH (2006) Detailed analysis of intrahepatic CD8 T cells in the normal and hepatitis C-infected liver reveals differences in specific populations of memory cells with distinct homing phenotypes. J Immunol 177(1): 729-738 Hirakawa S, Brown LF, Kodama S, Paavonen K, Alitalo K, Detmar M (2006) VEGF- C-induced lymphangiogenesis in sentinel lymph nodes promotes tumor metastasis to distant sites. Blood: [Epub ahead of print] Hirakawa S, Brown LF, Kodama S, Paavonen K, Alitalo K, Detmar M (2007) VEGF- C-induced lymphangiogenesis in sentinel lymph nodes promotes tumor metastasis to distant sites. Blood 109(3): 1010-1017 133 Hirakawa S, Fujii S, Kajiya K, Yano K, Detmar M (2005a) Vascular endothelial growth factor promotes sensitivity to ultraviolet B-induced cutaneous photodamage. Blood 105(6): 2392-2399 Hirakawa S, Hong YK, Harvey N, Schacht V, Matsuda K, Libermann T, Detmar M (2003) Identification of vascular lineage-specific genes by transcriptional profiling of isolated blood vascular and lymphatic endothelial cells. Am J Pathol 162(2): 575-586 Hirakawa S, Kodama S, Kunstfeld R, Kajiya K, Brown LF, Detmar M (2005b) VEGF-A induces tumor and sentinel lymph node lymphangiogenesis and promotes lymphatic metastasis. J Exp Med 201(7): 1089-1099 Holash J, Maisonpierre PC, Compton D, Boland P, Alexander CR, Zagzag D, Yancopoulos GD, Wiegand SJ (1999a) Vessel cooption, regression, and growth in tumors mediated by angiopoietins and VEGF. Science 284(5422): 1994-1998 Holash J, Wiegand SJ, Yancopoulos GD (1999b) New model of tumor angiogenesis: dynamic balance between vessel regression and growth mediated by angiopoietins and VEGF. Oncogene 18(38): 5356-5362 Hong YK, Detmar M (2003a) Prox1, master regulator of the lymphatic vasculature phenotype. Cell Tissue Res 314(1): 85-92 Hong YK, Detmar M (2003b) Prox1, master regulator of the lymphatic vasculature phenotype. Cell Tissue Res 314: 85-92 Hong YK, Foreman K, Shin JW, Hirakawa S, Curry CL, Sage DR, Libermann T, Dezube BJ, Fingeroth JD, Detmar M (2004a) Lymphatic reprogramming of blood vascular endothelium by Kaposi sarcoma-associated herpesvirus. Nat Genet 36(7): 683-685 Hong YK, Harvey N, Noh YH, Schacht V, Hirakawa S, Detmar M, Oliver G (2002) Prox1 is a master control gene in the program specifying lymphatic endothelial cell fate. Dev Dyn 225(3): 351-357 Hong YK, Lange-Asschenfeldt B, Velasco P, Hirakawa S, Kunstfeld R, Brown LF, Bohlen P, Senger DR, Detmar M (2004b) VEGF-A promotes tissue repair-associated lymphatic vessel formation via VEGFR-2 and the alpha1beta1 and alpha2beta1 integrins. Faseb J 18(10): 1111-1113 Hong YK, Shin JW, Detmar M (2004c) Development of the lymphatic vascular system: a mystery unravels. Dev Dyn 231(3): 462-473 Houck KA, Ferrara N, Winer J, Cachianes G, Li B, Leung DW (1991) The vascular endothelial growth factor family: identification of a fourth molecular species and characterization of alternative splicing of RNA. Mol Endocrinol 5(12): 1806-1814 Houghton AN, Albino AP, Cordon-Cardo C, Davis LJ, Eisinger M (1988) Cell surface antigens of human melanocytes and melanoma. Expression of adenosine 134 deaminase binding protein is extinguished with melanocyte transformation. J Exp Med 167(1): 197-212 Huber W, von Heydebreck A, Sultmann H, Poustka A, Vingron M (2002) Variance stabilization applied to microarray data calibration and to the quantification of differential expression. Bioinformatics 18 Suppl 1: S96-104 Huntington GS, McClure CFW (1910) The anatomy and development of the jugular lymph sac in the domestic cat (Felis domestica). Am J Anat 10: 177-311 Irrthum A, Devriendt K, Chitayat D, Matthijs G, Glade C, Steijlen PM, Fryns JP, Van Steensel MA, Vikkula M (2003) Mutations in the transcription factor gene SOX18 underlie recessive and dominant forms of hypotrichosis-lymphedema-telangiectasia. Am J Hum Genet 72(6): 1470-1478 Irrthum A, Karkkainen MJ, Devriendt K, Alitalo K, Vikkula M (2000) Congenital hereditary lymphedema caused by a mutation that inactivates VEGFR3 tyrosine kinase. Am J Hum Genet 67(2): 295-301 Jackson DG (2004) Biology of the lymphatic marker LYVE-1 and applications in research into lymphatic trafficking and lymphangiogenesis. Apmis 112(7-8): 526-538 Jackson DG, Prevo R, Clasper S, Banerji S (2001) LYVE-1, the lymphatic system and tumor lymphangiogenesis. Trends Immunol 22(6): 317-321 Javerzat S, Auguste P, Bikfalvi A (2002) The role of fibroblast growth factors in vascular development. Trends Mol Med 8(10): 483-489 Jeltsch M, Kaipainen A, Joukov V, Meng X, Lakso M, Rauvala H, Swartz M, Fukumura D, Jain RK, Alitalo K (1997) Hyperplasia of lymphatic vessels in VEGF-C transgenic mice. Science 276(5317): 1423-1425 Joukov V, Pajusola K, Kaipainen A, Chilov D, Lahtinen I, Kukk E, Saksela O, Kalkkinen N, Alitalo K (1996) A novel vascular endothelial growth factor, VEGF-C, is a ligand for the Flt4 (VEGFR-3) and KDR (VEGFR-2) receptor tyrosine kinases. Embo J 15(7): 1751 Joukov V, Sorsa T, Kumar V, Jeltsch M, Claesson-Welsh L, Cao Y, Saksela O, Kalkkinen N, Alitalo K (1997) Proteolytic processing regulates receptor specificity and activity of VEGF-C. Embo J 16(13): 3898-3911 Kaipainen A, Korhonen J, Mustonen T, van Hinsbergh VW, Fang GH, Dumont D, Breitman M, Alitalo K (1995) Expression of the fms-like tyrosine kinase 4 gene becomes restricted to lymphatic endothelium during development. Proc Natl Acad Sci U S A 92(8): 3566-3570 Kaipainen A, Korhonen J, Pajusola K, Aprelikova O, Persico MG, Terman BI, Alitalo K (1993) The related FLT4, FLT1, and KDR receptor tyrosine kinases show distinct expression patterns in human fetal endothelial cells. J Exp Med 178(6): 2077-2088 135 Kajiya K, Detmar M (2006) An important role of lymphatic vessels in the control of UVB-induced edema formation and inflammation. J Invest Dermatol 126(4): 919-921 Kajiya K, Hirakawa S, Ma B, Drinnenberg I, Detmar M (2005) Hepatocyte growth factor promotes lymphatic vessel formation and function. Embo J 24(16): 2885-2895 Kanai M, Goke M, Tsunekawa S, Podolsky DK (1997) Signal transduction pathway of human fibroblast growth factor receptor 3. Identification of a novel 66-kDa phosphoprotein. J Biol Chem 272(10): 6621-6628 Karkkainen MJ, Ferrell RE, Lawrence EC, Kimak MA, Levinson KL, McTigue MA, Alitalo K, Finegold DN (2000) Missense mutations interfere with VEGFR-3 signalling in primary lymphoedema. Nat Genet 25(2): 153-159 Karkkainen MJ, Haiko P, Sainio K, Partanen J, Taipale J, Petrova TV, Jeltsch M, Jackson DG, Talikka M, Rauvala H, Betsholtz C, Alitalo K (2004) Vascular endothelial growth factor C is required for sprouting of the first lymphatic vessels from embryonic veins. Nat Immunol 5(1): 74-80 Karkkainen MJ, Jussila L, Ferrell RE, Finegold DN, Alitalo K (2001) Molecular regulation of lymphangiogenesis and targets for tissue oedema. Trends Mol Med 7(1): 18-22 Karpanen T, Wirzenius M, Makinen T, Veikkola T, Haisma HJ, Achen MG, Stacker SA, Pytowski B, Yla-Herttuala S, Alitalo K (2006) Lymphangiogenic growth factor responsiveness is modulated by postnatal lymphatic vessel maturation. Am J Pathol 169(2): 708-718 Kedde M, Strasser MJ, Boldajipour B, Vrielink JA, Slanchev K, le Sage C, Nagel R, Voorhoeve PM, van Duijse J, Orom UA, Lund AH, Perrakis A, Raz E, Agami R (2007) RNA-binding protein Dnd1 inhibits microRNA access to target mRNA. Cell 131(7): 1273-1286 Kerjaschki D (2005) The crucial role of macrophages in lymphangiogenesis. J Clin Invest 115(9): 2316-2319 Kerjaschki D, Huttary N, Raab I, Regele H, Bojarski-Nagy K, Bartel G, Krober SM, Greinix H, Rosenmaier A, Karlhofer F, Wick N, Mazal PR (2006) Lymphatic endothelial progenitor cells contribute to de novo lymphangiogenesis in human renal transplants. Nat Med 12(2): 230-234 Kerjaschki D, Regele HM, Moosberger I, Nagy-Bojarski K, Watschinger B, Soleiman A, Birner P, Krieger S, Hovorka A, Silberhumer G, Laakkonen P, Petrova T, Langer B, Raab I (2004) Lymphatic neoangiogenesis in human kidney transplants is associated with immunologically active lymphocytic infiltrates. J Am Soc Nephrol 15(3): 603-612 Kersten S, Mandard S, Tan NS, Escher P, Metzger D, Chambon P, Gonzalez FJ, Desvergne B, Wahli W (2000) Characterization of the fasting-induced adipose factor 136 FIAF, a novel peroxisome proliferator-activated receptor target gene. J Biol Chem 275(37): 28488-28493 Kertesz Z, Linton EA, Redman CW (2000) Adhesion molecules of syncytiotrophoblast microvillous membranes inhibit proliferation of human umbilical vein endothelial cells. Placenta 21(2-3): 150-159 Kilroy GE, Foster SJ, Wu X, Ruiz J, Sherwood S, Heifetz A, Ludlow JW, Stricker DM, Potiny S, Green P, Halvorsen YD, Cheatham B, Storms RW, Gimble JM (2007) Cytokine profile of human adipose-derived stem cells: expression of angiogenic, hematopoietic, and pro-inflammatory factors. J Cell Physiol 212(3): 702-709 Kim I, Kim HG, Kim H, Kim HH, Park SK, Uhm CS, Lee ZH, Koh GY (2000) Hepatic expression, synthesis and secretion of a novel fibrinogen/angiopoietin-related protein that prevents endothelial-cell apoptosis. Biochem J 346 Pt 3: 603-610 Kimura H, Konishi K, Nukui T, Kaji M, Maeda K, Yabushita K, Tsuji M, Miwa A (2001) Prognostic significance of expression of thymidine phosphorylase and vascular endothelial growth factor in human gastric carcinoma. J Surg Oncol 76(1): 31-36 Kriederman BM, Myloyde TL, Witte MH, Dagenais SL, Witte CL, Rennels M, Bernas MJ, Lynch MT, Erickson RP, Caulder MS, Miura N, Jackson D, Brooks BP, Glover TW (2003) FOXC2 haploinsufficient mice are a model for human autosomal dominant lymphedema-distichiasis syndrome. Hum Mol Genet 12(10): 1179-1185 Kriehuber E, Breiteneder-Geleff S, Groeger M, Soleiman A, Schoppmann SF, Stingl G, Kerjaschki D, Maurer D (2001) Isolation and characterization of dermal lymphatic and blood endothelial cells reveal stable and functionally specialized cell lineages. J Exp Med 194(6): 797-808 Kubo H, Cao R, Brakenhielm E, Makinen T, Cao Y, Alitalo K (2002) Blockade of vascular endothelial growth factor receptor-3 signaling inhibits fibroblast growth factor-2-induced lymphangiogenesis in mouse cornea. Proc Natl Acad Sci U S A 99(13): 8868-8873 Kubo H, Fujiwara T, Jussila L, Hashi H, Ogawa M, Shimizu K, Awane M, Sakai Y, Takabayashi A, Alitalo K, Yamaoka Y, Nishikawa SI (2000) Involvement of vascular endothelial growth factor receptor-3 in maintenance of integrity of endothelial cell lining during tumor angiogenesis. Blood 96(2): 546-553 Kukk E, Lymboussaki A, Taira S, Kaipainen A, Jeltsch M, Joukov V, Alitalo K (1996) VEGF-C receptor binding and pattern of expression with VEGFR-3 suggests a role in lymphatic vascular development. Development 122(12): 3829-3837 Kunstfeld R, Hirakawa S, Hong YK, Schacht V, Lange-Asschenfeldt B, Velasco P, Lin C, Fiebiger E, Wei X, Wu Y, Hicklin D, Bohlen P, Detmar M (2004) Induction of cutaneous delayed-type hypersensitivity reactions in VEGF-A transgenic mice results in chronic skin inflammation associated with persistent lymphatic hyperplasia. Blood 104(4): 1048-1057 137 Kuo WP, Liu F, Trimarchi J, Punzo C, Lombardi M, Sarang J, Whipple ME, Maysuria M, Serikawa K, Lee SY, McCrann D, Kang J, Shearstone JR, Burke J, Park DJ, Wang X, Rector TL, Ricciardi-Castagnoli P, Perrin S, Choi S, Bumgarner R, Kim JH, Short GF, 3rd, Freeman MW, Seed B, Jensen R, Church GM, Hovig E, Cepko CL, Park P, Ohno-Machado L, Jenssen TK (2006) A sequence-oriented comparison of gene expression measurements across different hybridization-based technologies. Nat Biotechnol 24(7): 832-840 Lange T, Guttmann-Raviv N, Baruch L, Machluf M, Neufeld G (2003) VEGF162, a new heparin-binding vascular endothelial growth factor splice form that is expressed in transformed human cells. J Biol Chem 278(19): 17164-17169 Lengler J, Krausz E, Tomarev S, Prescott A, Quinlan RA, Graw J (2001) Antagonistic action of Six3 and Prox1 at the gamma-crystallin promoter. Nucleic Acids Res 29(2): 515-526 Leu AJ, Gretener SB, Enderlin S, Bruhlmann P, Michel BA, Kowal-Bielecka O, Hoffmann U, Franzeck UK (1999) Lymphatic microangiopathy of the skin in systemic sclerosis. Rheumatology (Oxford) 38(3): 221-227 Li JH, Kluger MS, Madge LA, Zheng L, Bothwell AL, Pober JS (2002) Interferon- gamma augments CD95(APO-1/Fas) and pro-caspase-8 expression and sensitizes human vascular endothelial cells to CD95-mediated apoptosis. Am J Pathol 161(4): 1485-1495 Liang Y, Kelemen A (2007) Bayesian state space models for inferring and predicting temporal gene expression profiles. Biom J 49(6): 801-814 Liu D, Jia H, Holmes DI, Stannard A, Zachary I (2003a) Vascular endothelial growth factor-regulated gene expression in endothelial cells: KDR-mediated induction of Egr3 and the related nuclear receptors Nur77, Nurr1, and Nor1. Arterioscler Thromb Vasc Biol 23(11): 2002-2007 Liu L, Ratner BD, Sage EH, Jiang S (2007) Endothelial cell migration on surface- density gradients of fibronectin, VEGF, or both proteins. Langmuir 23(22): 11168- 11173 Liu YW, Gao W, Teh HL, Tan JH, Chan WK (2003b) Prox1 is a novel coregulator of Ff1b and is involved in the embryonic development of the zebra fish interrenal primordium. Mol Cell Biol 23(20): 7243-7255 Liu Z, Xu J, Colvin JS, Ornitz DM (2002) Coordination of chondrogenesis and osteogenesis by fibroblast growth factor 18. Genes Dev 16(7): 859-869 Loster K, Zeilinger K, Schuppan D, Reutter W (1995) The cysteine-rich region of dipeptidyl peptidase IV (CD 26) is the collagen-binding site. Biochem Biophys Res Commun 217(1): 341-348 Lu D, Kussie P, Pytowski B, Persaud K, Bohlen P, Witte L, Zhu Z (2000) Identification of the residues in the extracellular region of KDR important for 138 interaction with vascular endothelial growth factor and neutralizing anti-KDR antibodies. J Biol Chem 275(19): 14321-14330 Magic Z, Radulovic S, Brankovic-Magic M (2007) cDNA microarrays: identification of gene signatures and their application in clinical practice. J BUON 12 Suppl 1: S39- 44 Maisonpierre PC, Suri C, Jones PF, Bartunkova S, Wiegand SJ, Radziejewski C, Compton D, McClain J, Aldrich TH, Papadopoulos N, Daly TJ, Davis S, Sato TN, Yancopoulos GD (1997) Angiopoietin-2, a natural antagonist for Tie2 that disrupts in vivo angiogenesis. Science 277(5322): 55-60 Makinen T, Adams RH, Bailey J, Lu Q, Ziemiecki A, Alitalo K, Klein R, Wilkinson GA (2005) PDZ interaction site in ephrinB2 is required for the remodeling of lymphatic vasculature. Genes Dev 19(3): 397-410 Makinen T, Veikkola T, Mustjoki S, Karpanen T, Catimel B, Nice EC, Wise L, Mercer A, Kowalski H, Kerjaschki D, Stacker SA, Achen MG, Alitalo K (2001) Isolated lymphatic endothelial cells transduce growth, survival and migratory signals via the VEGF-C/D receptor VEGFR-3. Embo J 20(17): 4762-4773 Mandriota SJ, Jussila L, Jeltsch M, Compagni A, Baetens D, Prevo R, Banerji S, Huarte J, Montesano R, Jackson DG, Orci L, Alitalo K, Christofori G, Pepper MS (2001) Vascular endothelial growth factor-C-mediated lymphangiogenesis promotes tumour metastasis. Embo J 20(4): 672-682 Mandriota SJ, Seghezzi G, Vassalli JD, Ferrara N, Wasi S, Mazzieri R, Mignatti P, Pepper MS (1995) Vascular endothelial growth factor increases urokinase receptor expression in vascular endothelial cells. J Biol Chem 270(17): 9709-9716 Marconcini L, Marchio S, Morbidelli L, Cartocci E, Albini A, Ziche M, Bussolino F, Oliviero S (1999) c-fos-induced growth factor/vascular endothelial growth factor D induces angiogenesis in vivo and in vitro. Proc Natl Acad Sci U S A 96(17): 9671- 9676 Martinez I, Lombardia L, Garcia-Barreno B, Dominguez O, Melero JA (2007) Distinct gene subsets are induced at different time points after human respiratory syncytial virus infection of A549 cells. J Gen Virol 88(Pt 2): 570-581 Maruyama K, Ii M, Cursiefen C, Jackson DG, Keino H, Tomita M, Van Rooijen N, Takenaka H, D'Amore PA, Stein-Streilein J, Losordo DW, Streilein JW (2005) Inflammation-induced lymphangiogenesis in the cornea arises from CD11b-positive macrophages. J Clin Invest 115(9): 2363-2372 Massberg S, Schaerli P, Knezevic-Maramica I, Kollnberger M, Tubo N, Moseman EA, Huff IV, Junt T, Wagers AJ, Mazo IB, von Andrian UH (2007) Immunosurveillance by hematopoietic progenitor cells trafficking through blood, lymph, and peripheral tissues. Cell 131(5): 994-1008 139 McEwen DG, Ornitz DM (1998) Regulation of the fibroblast growth factor receptor 3 promoter and intron I enhancer by Sp1 family transcription factors. J Biol Chem 273(9): 5349-5357 McLatchie LM, Fraser NJ, Main MJ, Wise A, Brown J, Thompson N, Solari R, Lee MG, Foord SM (1998) RAMPs regulate the transport and ligand specificity of the calcitonin-receptor-like receptor. Nature 393(6683): 333-339 Mentlein R (2004) Cell-surface peptidases. Int Rev Cytol 235: 165-213 Millington OR, Zinselmeyer BH, Brewer JM, Garside P, Rush CM (2007) Lymphocyte tracking and interactions in secondary lymphoid organs. Inflamm Res 56(10): 391-401 Moller P, Bohm M, Czarnetszki BM, Schadendorf D (1996) Interleukin-7. Biology and implications for dermatology. Exp Dermatol 5(3): 129-137 Moscatelli D (1987) High and low affinity binding sites for basic fibroblast growth factor on cultured cells: absence of a role for low affinity binding in the stimulation of plasminogen activator production by bovine capillary endothelial cells. J Cell Physiol 131(1): 123-130 Muller A, Homey B, Soto H, Ge N, Catron D, Buchanan ME, McClanahan T, Murphy E, Yuan W, Wagner SN, Barrera JL, Mohar A, Verastegui E, Zlotnik A (2001) Involvement of chemokine receptors in breast cancer metastasis. Nature 410(6824): 50-56 Murgue B, Tsunekawa S, Rosenberg I, deBeaumont M, Podolsky DK (1994) Identification of a novel variant form of fibroblast growth factor receptor 3 (FGFR3 IIIb) in human colonic epithelium. Cancer Res 54(19): 5206-5211 Murphy D (2002) Gene expression studies using microarrays: principles, problems, and prospects. Adv Physiol Educ 26(1-4): 256-270 Nagy JA, Vasile E, Feng D, Sundberg C, Brown LF, Manseau EJ, Dvorak AM, Dvorak HF (2002) VEGF-A induces angiogenesis, arteriogenesis, lymphangiogenesis, and vascular malformations. Cold Spring Harb Symp Quant Biol 67: 227-237 Neufeld G, Cohen T, Gengrinovitch S, Poltorak Z (1999) Vascular endothelial growth factor (VEGF) and its receptors. FASEB J 13(1): 9-22 Neufeld G, Tessler S, Gitay-Goren H, Cohen T, Levi BZ (1994) Vascular endothelial growth factor and its receptors. Prog Growth Factor Res 5(1): 89-97 Niederkorn JY, Peeler JS, Ross J, Callanan D (1989) The immunogenic privilege of corneal allografts. Reg Immunol 2(2): 117-124 140 Ohl L, Mohaupt M, Czeloth N, Hintzen G, Kiafard Z, Zwirner J, Blankenstein T, Henning G, Forster R (2004) CCR7 governs skin dendritic cell migration under inflammatory and steady-state conditions. Immunity 21(2): 279-288 Oliver G (2004) Lymphatic vasculature development. Nat Rev Immunol 4(1): 35-45 Oliver G, Detmar M (2002) The rediscovery of the lymphatic system: old and new insights into the development and biological function of the lymphatic vasculature. Genes Dev 16(7): 773-783 Oliver G, Harvey N (2002) A stepwise model of the development of lymphatic vasculature. Ann N Y Acad Sci 979: 159-165; discussion 188-196 Oliver G, Sosa-Pineda B, Geisendorf S, Spana EP, Doe CQ, Gruss P (1993a) Prox 1, a prospero-related homeobox gene expressed during mouse development. Mech Dev 44(1): 3-16 Oliver G, Sosa-Pineda B, Geisendorf S, Spana EP, Doe CQ, Gruss P (1993b) Prox 1, a prospero-related homeobox gene expressed during mouse development. Mech Dev 44(1): 3-16. Orlandini M, Marconcini L, Ferruzzi R, Oliviero S (1996) Identification of a c-fos- induced gene that is related to the platelet-derived growth factor/vascular endothelial growth factor family. Proc Natl Acad Sci U S A 93(21): 11675-11680 Ornitz DM (2000) FGFs, heparan sulfate and FGFRs: complex interactions essential for development. Bioessays 22(2): 108-112 Ornitz DM, Itoh N (2001) Fibroblast growth factors. Genome Biol 2(3): REVIEWS3005 Ornitz DM, Marie PJ (2002) FGF signaling pathways in endochondral and intramembranous bone development and human genetic disease. Genes Dev 16(12): 1446-1465 Ornitz DM, Xu J, Colvin JS, McEwen DG, MacArthur CA, Coulier F, Gao G, Goldfarb M (1996) Receptor specificity of the fibroblast growth factor family. J Biol Chem 271(25): 15292-15297 Ornitz DM, Yayon A, Flanagan JG, Svahn CM, Levi E, Leder P (1992) Heparin is required for cell-free binding of basic fibroblast growth factor to a soluble receptor and for mitogenesis in whole cells. Mol Cell Biol 12(1): 240-247 Orr-Urtreger A, Bedford MT, Burakova T, Arman E, Zimmer Y, Yayon A, Givol D, Lonai P (1993) Developmental localization of the splicing alternatives of fibroblast growth factor receptor-2 (FGFR2). Dev Biol 158(2): 475-486. Pan W (2002) A comparative review of statistical methods for discovering differentially expressed genes in replicated microarray experiments. Bioinformatics 18(4): 546-554 141 Patel DD, Koopmann W, Imai T, Whichard LP, Yoshie O, Krangel MS (2001) Chemokines have diverse abilities to form solid phase gradients. Clin Immunol 99(1): 43-52 Patterson TA, Lobenhofer EK, Fulmer-Smentek SB, Collins PJ, Chu TM, Bao W, Fang H, Kawasaki ES, Hager J, Tikhonova IR, Walker SJ, Zhang L, Hurban P, de Longueville F, Fuscoe JC, Tong W, Shi L, Wolfinger RD (2006) Performance comparison of one-color and two-color platforms within the MicroArray Quality Control (MAQC) project. Nat Biotechnol 24(9): 1140-1150 Pepper MS (2001) Lymphangiogenesis and tumor metastasis: myth or reality? Clin Cancer Res 7(3): 462-468 Perollet C, Han ZC, Savona C, Caen JP, Bikfalvi A (1998) Platelet factor 4 modulates fibroblast growth factor 2 (FGF-2) activity and inhibits FGF-2 dimerization. Blood 91(9): 3289-3299 Persaud K, Tille JC, Liu M, Zhu Z, Jimenez X, Pereira DS, Miao HQ, Brennan LA, Witte L, Pepper MS, Pytowski B (2004) Involvement of the VEGF receptor 3 in tubular morphogenesis demonstrated with a human anti-human VEGFR-3 monoclonal antibody that antagonizes receptor activation by VEGF-C. J Cell Sci 117(Pt 13): 2745-2756 Petrova TV, Karpanen T, Norrmen C, Mellor R, Tamakoshi T, Finegold D, Ferrell R, Kerjaschki D, Mortimer P, Yla-Herttuala S, Miura N, Alitalo K (2004) Defective valves and abnormal mural cell recruitment underlie lymphatic vascular failure in lymphedema distichiasis. Nat Med 10(9): 974-981 Petrova TV, Makinen T, Makela TP, Saarela J, Virtanen I, Ferrell RE, Finegold DN, Kerjaschki D, Yla-Herttuala S, Alitalo K (2002a) Lymphatic endothelial reprogramming of vascular endothelial cells by the Prox-1 homeobox transcription factor. Embo J 21(17): 4593-4599 Petrova TV, Makinen T, Makela TP, Saarela J, Virtanen I, Ferrell RE, Finegold DN, Kerjaschki D, Yla-Herttuala S, Alitalo K (2002b) Lymphatic endothelial reprogramming of vascular endothelial cells by the Prox-1 homeobox transcription factor. EMBO J 21(17): 4593-4599 Plouet J, Moro F, Bertagnolli S, Coldeboeuf N, Mazarguil H, Clamens S, Bayard F (1997) Extracellular cleavage of the vascular endothelial growth factor 189-amino acid form by urokinase is required for its mitogenic effect. J Biol Chem 272(20): 13390-13396 Podgrabinska S, Braun P, Velasco P, Kloos B, Pepper MS, Skobe M (2002) Molecular characterization of lymphatic endothelial cells. Proc Natl Acad Sci U S A 99(25): 16069-16074 Powers CJ, McLeskey SW, Wellstein A (2000) Fibroblast growth factors, their receptors and signaling. Endocr Relat Cancer 7(3): 165-197 142 Prabhakarpandian B, Goetz DJ, Swerlick RA, Chen X, Kiani MF (2001) Expression and functional significance of adhesion molecules on cultured endothelial cells in response to ionizing radiation. Microcirculation 8(5): 355-364 Prevo R, Banerji S, Ferguson DJ, Clasper S, Jackson DG (2001) Mouse LYVE-1 is an endocytic receptor for hyaluronan in lymphatic endothelium. J Biol Chem 276(22): 19420-19430 Proost P, Menten P, Struyf S, Schutyser E, De Meester I, Van Damme J (2000) Cleavage by CD26/dipeptidyl peptidase IV converts the chemokine LD78beta into a most efficient monocyte attractant and CCR1 agonist. Blood 96(5): 1674-1680 Proost P, Schutyser E, Menten P, Struyf S, Wuyts A, Opdenakker G, Detheux M, Parmentier M, Durinx C, Lambeir AM, Neyts J, Liekens S, Maudgal PC, Billiau A, Van Damme J (2001) Amino-terminal truncation of CXCR3 agonists impairs receptor signaling and lymphocyte chemotaxis, while preserving antiangiogenic properties. Blood 98(13): 3554-3561 Proost P, Struyf S, Schols D, Opdenakker G, Sozzani S, Allavena P, Mantovani A, Augustyns K, Bal G, Haemers A, Lambeir AM, Scharpe S, Van Damme J, De Meester I (1999) Truncation of macrophage-derived chemokine by CD26/ dipeptidyl- peptidase IV beyond its predicted cleavage site affects chemotactic activity and CC chemokine receptor 4 interaction. J Biol Chem 274(7): 3988-3993 Pugh CW, Ratcliffe PJ (2003) Regulation of angiogenesis by hypoxia: role of the HIF system. Nat Med 9(6): 677-684 Qian J, Lin J, Luscombe NM, Yu H, Gerstein M (2003) Prediction of regulatory networks: genome-wide identification of transcription factor targets from gene expression data. Bioinformatics 19(15): 1917-1926 Ramaswamy S, Ross KN, Lander ES, Golub TR (2003) A molecular signature of metastasis in primary solid tumors. Nat Genet 33(1): 49-54 Randolph GJ, Angeli V, Swartz MA (2005) Dendritic-cell trafficking to lymph nodes through lymphatic vessels. Nat Rev Immunol 5(8): 617-628 Rennel E, Mellberg S, Dimberg A, Petersson L, Botling J, Ameur A, Westholm JO, Komorowski J, Lassalle P, Cross MJ, Gerwins P (2007) Endocan is a VEGF-A and PI3K regulated gene with increased expression in human renal cancer. Exp Cell Res 313(7): 1285-1294 Richard L, Velasco P, Detmar M (1998) A simple immunomagnetic protocol for the selective isolation and long-term culture of human dermal microvascular endothelial cells. Exp Cell Res 240(1): 1-6 Rishi AK, Joyce-Brady M, Fisher J, Dobbs LG, Floros J, VanderSpek J, Brody JS, Williams MC (1995) Cloning, characterization, and development expression of a rat 143 lung alveolar type I cell gene in embryonic endodermal and neural derivatives. Dev Biol 167(1): 294-306 Robinson CJ, Stringer SE (2001) The splice variants of vascular endothelial growth factor (VEGF) and their receptors. J Cell Sci 114(Pt 5): 853-865 Rockson SG (2001) Lymphedema. Am J Med 110(4): 288-295 Roesli C, Mumprecht V, Neri D, Detmar M (2008) Identification of the surface- accessible, lineage-specific vascular proteome by two-dimensional peptide mapping. Faseb J Rossiter H, Barresi C, Pammer J, Rendl M, Haigh J, Wagner EF, Tschachler E (2004) Loss of vascular endothelial growth factor a activity in murine epidermal keratinocytes delays wound healing and inhibits tumor formation. Cancer Res 64(10): 3508-3516 Roth V (2004) The generalized LASSO. IEEE Trans Neural Netw 15(1): 16-28 Ryter JM, Doe CQ, Matthews BW (2002) Structure of the DNA binding region of prospero reveals a novel homeo-prospero domain. Structure (Camb) 10(11): 1541- 1549 Saaristo A, Veikkola T, Enholm B, Hytonen M, Arola J, Pajusola K, Turunen P, Jeltsch M, Karkkainen MJ, Kerjaschki D, Bueler H, Yla-Herttuala S, Alitalo K (2002) Adenoviral VEGF-C overexpression induces blood vessel enlargement, tortuosity, and leakiness but no sprouting angiogenesis in the skin or mucous membranes. Faseb J 16(9): 1041-1049 Sabin FR (1902) On the origin of the lymphatic system from the veins and the development of the lymph hearts and thoracic duct in the pig. Am J Anat 1: 367-391 Saharinen P, Tammela T, Karkkainen MJ, Alitalo K (2004) Lymphatic vasculature: development, molecular regulation and role in tumor metastasis and inflammation. Trends Immunol 25(7): 387-395 Schacht V, Dadras SS, Johnson LA, Jackson DG, Hong YK, Detmar M (2005) Up- regulation of the lymphatic marker podoplanin, a mucin-type transmembrane glycoprotein, in human squamous cell carcinomas and germ cell tumors. Am J Pathol 166(3): 913-921 Schacht V, Ramirez MI, Hong YK, Hirakawa S, Feng D, Harvey N, Williams M, Dvorak AM, Dvorak HF, Oliver G, Detmar M (2003) T1alpha/podoplanin deficiency disrupts normal lymphatic vasculature formation and causes lymphedema. Embo J 22(14): 3546-3556 Schena M, Shalon D, Davis RW, Brown PO (1995) Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270(5235): 467- 470 144 Scherpereel A, Gentina T, Grigoriu B, Senechal S, Janin A, Tsicopoulos A, Plenat F, Bechard D, Tonnel AB, Lassalle P (2003) Overexpression of endocan induces tumor formation. Cancer Res 63(18): 6084-6089 Schledzewski K, Falkowski M, Moldenhauer G, Metharom P, Kzhyshkowska J, Ganss R, Demory A, Falkowska-Hansen B, Kurzen H, Ugurel S, Geginat G, Arnold B, Goerdt S (2006) Lymphatic endothelium-specific hyaluronan receptor LYVE-1 is expressed by stabilin-1+, F4/80+, CD11b+ macrophages in malignant tumours and wound healing tissue in vivo and in bone marrow cultures in vitro: implications for the assessment of lymphangiogenesis. J Pathol 209(1): 67-77 Schoenfeld J, Lessan K, Johnson NA, Charnock-Jones DS, Evans A, Vourvouhaki E, Scott L, Stephens R, Freeman TC, Saidi SA, Tom B, Weston GC, Rogers P, Smith SK, Print CG (2004) Bioinformatic analysis of primary endothelial cell gene array data illustrated by the analysis of transcriptome changes in endothelial cells exposed to VEGF-A and PlGF. Angiogenesis 7(2): 143-156 Schoppmann SF, Birner P, Stockl J, Kalt R, Ullrich R, Caucig C, Kriehuber E, Nagy K, Alitalo K, Kerjaschki D (2002) Tumor-associated macrophages express lymphatic endothelial growth factors and are related to peritumoral lymphangiogenesis. Am J Pathol 161(3): 947-956 Senger DR, Galli SJ, Dvorak AM, Perruzzi CA, Harvey VS, Dvorak HF (1983) Tumor cells secrete a vascular permeability factor that promotes accumulation of ascites fluid. Science 219(4587): 983-985 Shalon D, Smith SJ, Brown PO (1996) A DNA microarray system for analyzing complex DNA samples using two-color fluorescent probe hybridization. Genome Res 6(7): 639-645 Shields JD, Fleury ME, Yong C, Tomei AA, Randolph GJ, Swartz MA (2007) Autologous chemotaxis as a mechanism of tumor cell homing to lymphatics via interstitial flow and autocrine CCR7 signaling. Cancer Cell 11(6): 526-538 Shin JW, Min M, Larrieu-Lahargue F, Canron X, Kunstfeld R, Nguyen L, Henderson JE, Bikfalvi A, Detmar M, Hong YK (2006) Prox1 promotes lineage-specific expression of fibroblast growth factor (FGF) receptor-3 in lymphatic endothelium: a role for FGF signaling in lymphangiogenesis. Mol Biol Cell 17(2): 576-584 Shing Y, Folkman J, Sullivan R, Butterfield C, Murray J, Klagsbrun M (1984) Heparin affinity: purification of a tumor-derived capillary endothelial cell growth factor. Science 223(4642): 1296-1299 Skobe M, Hawighorst T, Jackson DG, Prevo R, Janes L, Velasco P, Riccardi L, Alitalo K, Claffey K, Detmar M (2001) Induction of tumor lymphangiogenesis by VEGF-C promotes breast cancer metastasis. Nat Med 7(2): 192-198 Sokal RR (1965) Statistical Methods in Systematics. Biol Rev Camb Philos Soc 40: 337-391 145 Sosa-Pineda B, Wigle JT, Oliver G (2000) Hepatocyte migration during liver development requires Prox1. Nat Genet 25(3): 254-255 Spira A, Beane JE, Shah V, Steiling K, Liu G, Schembri F, Gilman S, Dumas YM, Calner P, Sebastiani P, Sridhar S, Beamis J, Lamb C, Anderson T, Gerry N, Keane J, Lenburg ME, Brody JS (2007) Airway epithelial gene expression in the diagnostic evaluation of smokers with suspect lung cancer. Nat Med 13(3): 361-366 Stacker SA, Caesar C, Baldwin ME, Thornton GE, Williams RA, Prevo R, Jackson DG, Nishikawa S, Kubo H, Achen MG (2001) VEGF-D promotes the metastatic spread of tumor cells via the lymphatics. Nat Med 7(2): 186-191 Stacker SA, Stenvers K, Caesar C, Vitali A, Domagala T, Nice E, Roufail S, Simpson RJ, Moritz R, Karpanen T, Alitalo K, Achen MG (1999) Biosynthesis of vascular endothelial growth factor-D involves proteolytic processing which generates non- covalent homodimers. J Biol Chem 274(45): 32127-32136 Stalmans I, Ng YS, Rohan R, Fruttiger M, Bouche A, Yuce A, Fujisawa H, Hermans B, Shani M, Jansen S, Hicklin D, Anderson DJ, Gardiner T, Hammes HP, Moons L, Dewerchin M, Collen D, Carmeliet P, D'Amore PA (2002) Arteriolar and venular patterning in retinas of mice selectively expressing VEGF isoforms. J Clin Invest 109(3): 327-336 Storey JD, Xiao W, Leek JT, Tompkins RG, Davis RW (2005) Significance analysis of time course microarray experiments. Proc Natl Acad Sci U S A 102(36): 12837- 12842 Storkebaum E, Lambrechts D, Carmeliet P (2004) VEGF: once regarded as a specific angiogenic factor, now implicated in neuroprotection. Bioessays 26(9): 943-954 Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, Paulovich A, Pomeroy SL, Golub TR, Lander ES, Mesirov JP (2005) Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 102(43): 15545-15550 Suhardja A, Hoffman H (2003) Role of growth factors and their receptors in proliferation of microvascular endothelial cells. Microsc Res Tech 60(1): 70-75 Suri C, Jones PF, Patan S, Bartunkova S, Maisonpierre PC, Davis S, Sato TN, Yancopoulos GD (1996) Requisite role of angiopoietin-1, a ligand for the TIE2 receptor, during embryonic angiogenesis. Cell 87(7): 1171-1180 Swartz MA (2001) The physiology of the lymphatic system. Adv Drug Deliv Rev 50(1-2): 3-20 Tammela T, Saaristo A, Holopainen T, Lyytikka J, Kotronen A, Pitkonen M, Abo- Ramadan U, Yla-Herttuala S, Petrova TV, Alitalo K (2007) Therapeutic differentiation and maturation of lymphatic vessels after lymph node dissection and transplantation. Nat Med 13(12): 1458-1466 146 Tangemann K, Gunn MD, Giblin P, Rosen SD (1998) A high endothelial cell-derived chemokine induces rapid, efficient, and subset-selective arrest of rolling T lymphocytes on a reconstituted endothelial substrate. J Immunol 161(11): 6330-6337 Taniguchi K, Kohno R, Ayada T, Kato R, Ichiyama K, Morisada T, Oike Y, Yonemitsu Y, Maehara Y, Yoshimura A (2007) Spreds are essential for embryonic lymphangiogenesis by regulating vascular endothelial growth factor receptor 3 signaling. Mol Cell Biol 27(12): 4541-4550 Terada M, Shimizu A, Sato N, Miyakaze SI, Katayama H, Kurokawa-Seo M (2001) Fibroblast growth factor receptor 3 lacking the Ig IIIb and transmembrane domains secreted from human squamous cell carcinoma DJM-1 binds to FGFs. Mol Cell Biol Res Commun 4(6): 365-373 Thurston G (2003) Role of Angiopoietins and Tie receptor tyrosine kinases in angiogenesis and lymphangiogenesis. Cell Tissue Res 314(1): 61-68 Tischer E, Mitchell R, Hartman T, Silva M, Gospodarowicz D, Fiddes JC, Abraham JA (1991) The human gene for vascular endothelial growth factor. Multiple protein forms are encoded through alternative exon splicing. J Biol Chem 266(18): 11947- 11954 Tobler NE, Detmar M (2006) Tumor and lymph node lymphangiogenesis--impact on cancer metastasis. J Leukoc Biol 80(4): 691-696 Tomarev SI, Sundin O, Banerjee-Basu S, Duncan MK, Yang JM, Piatigorsky J (1996) Chicken homeobox gene Prox 1 related to Drosophila prospero is expressed in the developing lens and retina. Dev Dyn 206(4): 354-367 Unemori EN, Ferrara N, Bauer EA, Amento EP (1992) Vascular endothelial growth factor induces interstitial collagenase expression in human endothelial cells. J Cell Physiol 153(3): 557-562 Unger RE, Krump-Konvalinkova V, Peters K, Kirkpatrick CJ (2002) In vitro expression of the endothelial phenotype: comparative study of primary isolated cells and cell lines, including the novel cell line HPMEC-ST1.6R. Microvasc Res 64(3): 384-397 Vardhanabhuti S, Blakemore SJ, Clark SM, Ghosh S, Stephens RJ, Rajagopalan D (2006) A comparison of statistical tests for detecting differential expression using Affymetrix oligonucleotide microarrays. OMICS 10(4): 555-566 Veikkola T, Jussila L, Makinen T, Karpanen T, Jeltsch M, Petrova TV, Kubo H, Thurston G, McDonald DM, Achen MG, Stacker SA, Alitalo K (2001) Signalling via vascular endothelial growth factor receptor-3 is sufficient for lymphangiogenesis in transgenic mice. EMBO J 20(6): 1223-1231 Wei C, Li J, Bumgarner RE (2004) Sample size for detecting differentially expressed genes in microarray experiments. BMC Genomics 5(1): 87 147 Weinstein M, Xu X, Ohyama K, Deng CX (1998) FGFR-3 and FGFR-4 function cooperatively to direct alveogenesis in the murine lung. Development 125(18): 3615- 3623 Wesley UV, Albino AP, Tiwari S, Houghton AN (1999) A role for dipeptidyl peptidase IV in suppressing the malignant phenotype of melanocytic cells. J Exp Med 190(3): 311-322 Wetterwald A, Hoffstetter W, Cecchini MG, Lanske B, Wagner C, Fleisch H, Atkinson M (1996) Characterization and cloning of the E11 antigen, a marker expressed by rat osteoblasts and osteocytes. Bone 18(2): 125-132 Whitehurst B, Flister MJ, Bagaitkar J, Volk L, Bivens CM, Pickett B, Castro-Rivera E, Brekken RA, Gerard RD, Ran S (2007) Anti-VEGF-A therapy reduces lymphatic vessel density and expression of VEGFR-3 in an orthotopic breast tumor model. Int J Cancer 121(10): 2181-2191 Wigle JT, Harvey N, Detmar M, Lagutina I, Grosveld G, Gunn MD, Jackson DG, Oliver G (2002a) An essential role for Prox1 in the induction of the lymphatic endothelial cell phenotype. Embo J 21(7): 1505-1513 Wigle JT, Harvey N, Detmar M, Lagutina I, Grosveld G, Gunn MD, Jackson DG, Oliver G (2002b) An essential role for Prox1 in the induction of the lymphatic endothelial cell phenotype. Embo J 21(7): 1505-1513. Wigle JT, Oliver G (1999) Prox1 function is required for the development of the murine lymphatic system. Cell 98(6): 769-778. Wiley HE, Gonzalez EB, Maki W, Wu MT, Hwang ST (2001) Expression of CC chemokine receptor-7 and regional lymph node metastasis of B16 murine melanoma. J Natl Cancer Inst 93(21): 1638-1643 Wilkie AO, Patey SJ, Kan SH, van den Ouweland AM, Hamel BC (2002) FGFs, their receptors, and human limb malformations: clinical and molecular correlations. Am J Med Genet 112(3): 266-278 Withington ET (1894) Medical history from the earliest times, London: The Scientific Press. Witte MH, Erickson R, Bernas M, Andrade M, Reiser F, Conlon W, Hoyme HE, Witte CL (1998) Phenotypic and genotypic heterogeneity in familial Milroy lymphedema. Lymphology 31(4): 145-155 Wu LW, Mayo LD, Dunbar JD, Kessler KM, Baerwald MR, Jaffe EA, Wang D, Warren RS, Donner DB (2000) Utilization of distinct signaling pathways by receptors for vascular endothelial cell growth factor and other mitogens in the induction of endothelial cell proliferation. J Biol Chem 275(7): 5096-5103 148 Xia YP, Li B, Hylton D, Detmar M, Yancopoulos GD, Rudge JS (2003) Transgenic delivery of VEGF to mouse skin leads to an inflammatory condition resembling human psoriasis. Blood 102(1): 161-168 Xu J, Lawshe A, MacArthur CA, Ornitz DM (1999) Genomic structure, mapping, activity and expression of fibroblast growth factor 17. Mech Dev 83(1-2): 165-178 Xu J, Liu Z, Ornitz DM (2000) Temporal and spatial gradients of Fgf8 and Fgf17 regulate proliferation and differentiation of midline cerebellar structures. Development 127(9): 1833-1843 Yamada Y, Nezu J, Shimane M, Hirata Y (1997) Molecular cloning of a novel vascular endothelial growth factor, VEGF-D. Genomics 42(3): 483-488 Yan G, Fukabori Y, McBride G, Nikolaropolous S, McKeehan WL (1993) Exon switching and activation of stromal and embryonic fibroblast growth factor (FGF)- FGF receptor genes in prostate epithelial cells accompany stromal independence and malignancy. Mol Cell Biol 13(8): 4513-4522. Yanagi S, Inatome R, Takano T, Yamamura H (2001) Syk expression and novel function in a wide variety of tissues. Biochem Biophys Res Commun 288(3): 495-498 Yang S, Toy K, Ingle G, Zlot C, Williams PM, Fuh G, Li B, de Vos A, Gerritsen ME (2002) Vascular endothelial growth factor-induced genes in human umbilical vein endothelial cells: relative roles of KDR and Flt-1 receptors. Arterioscler Thromb Vasc Biol 22(11): 1797-1803 Yu SL, Chen HY, Chang GC, Chen CY, Chen HW, Singh S, Cheng CL, Yu CJ, Lee YC, Chen HS, Su TJ, Chiang CC, Li HN, Hong QS, Su HY, Chen CC, Chen WJ, Liu CC, Chan WK, Li KC, Chen JJ, Yang PC (2008) MicroRNA signature predicts survival and relapse in lung cancer. Cancer Cell 13(1): 48-57 Yuan L, Moyon D, Pardanaud L, Breant C, Karkkainen MJ, Alitalo K, Eichmann A (2002) Abnormal lymphatic vessel development in neuropilin 2 mutant mice. Development 129(20): 4797-4806 Zhang D, Zhang M (2007) Bayesian profiling of molecular signatures to predict event times. Theor Biol Med Model 4: 3 Zhang G, Fahmy RG, diGirolamo N, Khachigian LM (2006) JUN siRNA regulates matrix metalloproteinase-2 expression, microvascular endothelial growth and retinal neovascularisation. J Cell Sci 119(Pt 15): 3219-3226 Zukowska Z, Grant DS, Lee EW (2003) Neuropeptide Y: a novel mechanism for ischemic angiogenesis. Trends Cardiovasc Med 13(2): 86-92 149 7 APPENDIX 150 Appendix Table 1 LEC signature genes (sorted by median of AB1700 fold change) confirmed by LD-MDA LEC gene signature (236) Fold change by AB1700 Fold change by qPCR Gene name AB Sample Sample Sample Sample Sample Sample LD- Core probe ID 1 2 3 1 2 3 MDA guanylate cyclase 1, soluble, alpha 3 170165 497.3 155.0 83.6 1152.2 139.2 60.6 Yes deoxyribonuclease I-like 3 167226 49.3 2.6 51.3 peroxisome proliferative activated receptor, gamma 192239 37.8 60.2 10.1 49.8 58.6 8.6 Yes ATP-binding cassette, sub-family A (ABC1), member 4 194955 35.5 26.3 7.2 ADAMTS-like 3 158085 44.0 2.2 25.4 dipeptidylpeptidase 4 (CD26) 209451 35.9 21.6 12.1 70.7 13.8 9.1 Yes coxsackie virus and adenovirus receptor 108284 23.6 20.9 2.8 47.0 28.4 2.4 Yes doublecortin and CaM kinase-like 1 202776 20.0 10.4 28.6 41.5 13.6 15.8 Yes Yes interleukin 7 127208 34.6 2.6 17.8 93.9 5.5 24.1 Yes Yes collectin sub-family member 12 114422 39.3 17.2 6.1 114.1 21.9 7.1 Yes retinol binding protein 7, cellular 129740 16.8 1.4 29.6 426.9 161.5 512.2 Yes Yes phosphodiesterase 9A 105408 15.8 16.2 9.4 follistatin 117261 3.7 15.5 43.6 10.5 17.9 49.9 Yes runt-related transcription factor 1; translocated to, 1 122830 24.7 4.9 15.3 mannose receptor, C type 1 198568 119.4 6.4 15.3 250.3 7.6 9.0 Yes Yes cholesterol 25-hydroxylase 117883 130.4 3.6 15.3 pyruvate dehydrogenase kinase, isoenzyme 4 101060 175.1 12.5 14.7 RNA binding motif protein 35B 167987 41.1 10.2 14.6 ras homolog gene family, member U 130933 26.2 4.4 14.1 retinol binding protein 1, cellular 149921 70.0 13.9 7.5 209.6 30.5 12.7 Yes integrin, beta 4 184548 14.9 3.0 13.8 5222.0 375.0 530.4 Yes GRINL1A complex upstream protein 104996 193.2 13.8 9.0 chromosome 2 open reading frame 23 156624 68.7 13.5 7.8 periplakin 138890 16.8 12.5 12.8 12812.4 591.8 35.2 Yes Yes trefoil factor 3 (intestinal) 114445 44.8 8.1 12.7 460.9 14.8 19.5 Yes Yes DKFZP586A0522 protein 107957 64.7 4.0 12.6 solute carrier family 38, member 4 217080 12.5 5.3 17.0 141.5 10.1 21.2 Yes Yes carcinoembryonic antigen-related cell adhesion 219223 50.1 11.7 3.9 35.4 7.0 3.6 Yes molecule 1 growth hormone receptor 190306 13.1 9.0 11.7 33.9 18.2 9.5 Yes Yes IQ motif containing with AAA domain 152027 46.9 11.5 4.4 homeo box D10 166056 17.7 3.3 11.4 39.9 5.2 8.3 Yes Yes PDZ domain protein GIPC2 180419 20.8 5.5 11.1 64.1 8.1 20.1 Yes v-maf musculoaponeurotic fibrosarcoma oncogene 186589 28.4 10.7 4.2 61.9 12.7 3.7 Yes homolog chromosome 6 open reading frame 123 105756 69.4 10.6 7.4 relaxin 1 122881 11.8 10.6 2.2 CD36 antigen 121773 49.9 4.4 10.6 storkhead box 2 199062 9.9 3.2 16.7 aldehyde dehydrogenase 1 family, member A1 162248 10.1 1.8 9.6 34.2 2.4 13.8 Yes reelin 207609 20.1 8.2 9.6 40.4 9.5 7.8 Yes Yes prospero-related homeobox 1 124383 38.9 6.8 9.6 134.0 7.9 10.3 Yes Yes transient receptor potential cation channel C 6 101144 32.1 3.1 9.4 151 LEC gene signature (236) Fold change by AB1700 Fold change by qPCR Gene name AB Sample Sample Sample Sample Sample Sample LD- Core probe ID 1 2 3 1 2 3 MDA pleckstrin homology domain containing, family A 116372 8.5 13.7 9.3 member 2 phospholipase C, epsilon 1 132759 9.3 21.7 9.2 40.3 18.1 119.1 Yes Yes endothelin receptor type B 150558 83.9 6.6 9.2 150.8 7.9 15.7 Yes heparan sulfate (glucosamine) 3-O-sulfotransferase 1 154628 190.0 9.1 2.5 cadherin, EGF LAG seven-pass G-type receptor 1 202051 8.8 10.8 3.4 21.1 8.4 2.5 Yes hairy/enhancer-of-split related with YRPW motif 1 121998 8.5 2.5 8.6 20.3 5.7 9.8 Yes tissue inhibitor of metalloproteinase 3 179538 10.8 8.4 4.3 31.3 11.1 4.5 Yes Yes solute carrier family 26, member 4 194103 8.5 4.8 8.3 4.8 7.1 11.1 Yes zinc finger protein 467 184463 36.8 8.3 2.8 selenoprotein P, plasma, 1 169984 49.0 6.4 8.2 86.0 10.3 7.8 Yes Yes START domain containing 8 145078 8.1 2.1 11.1 GDNF family receptor alpha 1 202672 7.9 8.0 106.1 7.6 13.7 42.3 Yes Yes ring finger protein 152 217858 7.9 2.4 10.7 chemokine (C-C motif) ligand 21 138321 7.8 1.0 7.7 5259.7 86.8 42.7 Yes leucine rich repeat containing 1 115313 14.0 4.0 7.5 chromosome 8 open reading frame 55 143619 7.7 7.5 4.6 calmegin 100646 7.3 30.7 2.4 tumor necrosis factor receptor superfamily, member 11a 105151 11.8 7.3 5.9 LIM domain binding 2 127667 24.3 2.6 7.2 hyaluronoglucosaminidase 1 184118 75.2 5.8 7.2 682.7 9.0 20.2 Yes Yes SRY (sex determining region Y)-box 18 126548 26.9 7.2 4.3 98.9 7.4 9.0 Yes glia maturation factor, gamma 180184 36.6 2.9 7.1 ovostatin-2 138125 27.0 5.3 7.1 tumor necrosis factor (ligand) superfamily, member 10 153282 25.9 6.9 2.6 homeo box D4 219817 27.9 6.8 4.9 centrosomal protein 1 203266 6.7 2.8 7.1 cytochrome P450, family 1, subfamily A, polypeptide 1 135086 39.0 3.2 6.5 galanin 167511 10.9 2.3 6.5 klotho beta 120728 6.4 4.6 10.5 chromosome 10 open reading frame 116 121402 22.9 4.5 6.4 GTPase, IMAP family member 5 177981 97.7 6.2 5.6 chromosome 1 open reading frame 34 132797 19.6 2.1 6.0 coagulation factor C homolog, cochlin (Limulus 189694 7.7 4.7 5.9 polyphemus) lamin B receptor 159112 8.7 4.3 5.9 chromatin modifying protein 4C 177868 7.6 5.8 5.3 NDRG family member 2 153476 21.4 5.8 3.5 transmembrane protein 88 200951 40.6 3.5 5.6 sulfotransferase family, cytosolic, 1C, member 2 130755 21.5 5.5 5.0 leptin receptor 222462 5.5 3.0 6.6 13.6 8.5 9.1 Yes Yes serine protease inhibitor, Kunitz type, 2 212742 7.8 3.7 5.4 chromosome 17 open reading frame 28 115291 29.5 5.4 2.5 hydroxysteroid (17-beta) dehydrogenase 2 195707 21.5 2.7 5.4 39.9 4.0 9.1 Yes Yes phosphatidic acid phosphatase type 2B 133742 5.4 5.8 5.4 10.4 6.0 4.3 Yes DEAD (Asp-Glu-Ala-Asp) box polypeptide 10 166102 6.8 3.4 5.4 kinesin family member 14 121673 2.2 5.4 6.1 v-rel reticuloendotheliosis viral oncogene homolog 109539 6.1 5.3 2.4 (avian) integrin, alpha 9 157432 16.7 2.9 5.3 78.6 6.8 2.0 Yes Yes 152 LEC gene signature (236) Fold change by AB1700 Fold change by qPCR Gene name AB Sample Sample Sample Sample Sample Sample LD- Core probe ID 1 2 3 1 2 3 MDA fms-related tyrosine kinase 4 204569 19.9 3.0 5.1 83.6 6.6 5.9 Yes SH2 domain containing 3A 130169 10.7 3.0 5.1 YTH domain containing 2 181510 6.0 5.1 2.4 myosin VIIA and Rab interacting protein 155197 5.7 5.1 2.5 solute carrier family 24, member 1 216954 5.1 6.0 4.5 chondroitin beta1,4 N-acetylgalactosaminyltransferase 101140 10.0 5.0 3.3 23.1 6.6 4.1 Yes ring finger protein 144 101012 5.0 7.4 2.3 podoplanin 219722 30.8 2.5 5.0 80.5 3.6 7.2 Yes Yes ST6...N-acetylgalactosaminide alpha-2,6- 189728 55.5 3.8 5.0 sialyltransferase 3 chromosome 11 open reading frame 8 108279 5.0 3.0 31.9 dedicator of cytokinesis 8 106981 8.1 4.9 2.1 forkhead box C1 195499 4.0 4.9 4.8 8.7 5.5 3.5 Yes kelch repeat and BTB (POZ) domain containing 11 208898 9.4 4.1 4.8 fibroblast growth factor 12 138024 8.7 4.8 2.8 12.7 10.7 2.4 Yes proprotein convertase subtilisin/kexin type 6 154864 32.9 3.8 4.7 chromosome 18 open reading frame 30 171508 28.3 2.8 4.7 MADS box transcription enhancer factor 2, polypeptide 179021 12.3 4.7 1.8 30.1 5.6 5.9 Yes C protein phosphatase 1, regulatory (inhibitor) subunit 9A 198246 11.2 4.6 2.0 SH3 domain and tetratricopeptide repeats 2 133008 10.5 2.8 4.5 chromosome 7 open reading frame 29 138101 5.7 3.9 4.5 multimerin 1 128509 27.1 4.5 3.2 CTD (carboxy-terminal domain) small phosphatase-like 130122 4.5 4.6 3.6 zinc finger protein 650 211325 4.4 4.9 2.6 dehydrogenase/reductase (SDR family) member 3 125937 9.6 4.4 2.8 phosphodiesterase 6B, cGMP-specific, rod, beta 120762 5.9 4.3 2.2 toll-like receptor 4 216730 6.2 4.3 2.5 guanylate cyclase 1, soluble, beta 3 131372 4.3 7.8 3.9 secreted frizzled-related protein 1 143998 2.6 6.2 4.3 midline 1 (Opitz/BBB syndrome) 214412 5.2 2.7 4.2 SMAD, mothers against DPP homolog 9 (Drosophila) 106362 6.4 2.7 4.2 ADP-ribosylation factor-like 4B 146573 14.1 4.2 3.1 chromosome 2 open reading frame 31 204611 4.2 4.0 4.2 chromosome 20 open reading frame 129 199617 4.1 3.2 13.6 adrenergic, beta-2-, receptor, surface 127856 5.9 3.2 4.1 phosphoglucomutase 5 186266 6.8 4.1 3.3 v-yes-1 Yamaguchi sarcoma viral related oncogene 194134 6.6 4.1 3.9 12.9 4.2 3.5 Yes homolog B cell RAG associated protein 110602 5.5 4.0 2.0 neuritin 1 182524 8.7 4.0 3.2 SET and MYND domain containing 2 141023 4.0 4.0 5.4 ephrin-A5 158422 23.5 4.0 2.8 epithelial membrane protein 2 173909 4.0 3.5 7.4 adducin 3 (gamma) 132668 11.9 3.9 1.7 38.6 10.6 23.6 Yes Yes cell division cycle 25B 208879 4.0 2.8 3.9 serologically defined colon cancer antigen 33 186894 3.9 2.9 4.4 similar to tumor-associated membrane protein XMP 230120 3.5 3.9 4.7 transforming growth factor, alpha 180395 18.8 3.3 3.9 huntingtin-associated protein interacting protein 125617 9.3 3.7 3.9 153 LEC gene signature (236) Fold change by AB1700 Fold change by qPCR Gene name AB Sample Sample Sample Sample Sample Sample LD- Core probe ID 1 2 3 1 2 3 MDA piccolo (presynaptic cytomatrix protein) 176448 3.8 4.6 2.6 aquaporin 7 pseudogene 1 108084 19.3 3.8 3.2 galactosamine... N-acetylgalactosaminyltransferase-like 176965 7.5 3.8 2.7 4 suppressor of fused homolog (Drosophila) 176456 3.8 2.5 3.8 kelch-like 3 (Drosophila) 184006 3.1 3.8 4.7 cholesteryl ester transfer protein, plasma 140569 45.2 3.7 2.8 tetraspanin 12 215171 10.2 2.4 3.6 hydroxysteroid (17-beta) dehydrogenase 8 213046 3.3 3.6 3.8 homeo box D8 217040 7.6 3.6 3.6 sex comb on midleg-like 2 (Drosophila) 115809 5.5 3.3 3.6 tribbles homolog 2 (Drosophila) 188922 2.4 3.6 4.1 protein kinase C, zeta 184015 8.1 3.1 3.5 LIM domain only 2 (rhombotin-like 1) 207685 20.9 3.5 3.3 48.3 4.5 3.5 Yes nitric oxide synthase 3 (endothelial cell) 197614 14.6 3.3 3.5 chaperone, ABC1 activity of bc1 complex like (S. 144833 3.6 3.3 3.5 pombe) Rap guanine nucleotide exchange factor (GEF) 5 129115 6.9 3.4 2.8 glucose-fructose oxidoreductase domain containing 1 137251 2.8 3.4 3.7 ribosomal protein S6 kinase, 90kDa, polypeptide 5 152363 3.4 2.3 3.7 transmembrane 4 L six family member 18 125309 19.0 2.3 3.4 RALBP1 associated Eps domain containing 2 125119 15.9 3.4 2.2 arrestin, beta 1 111374 9.2 2.8 3.3 tissue factor pathway inhibitor 126759 5.2 2.2 3.3 aldehyde dehydrogenase 6 family, member A1 184174 3.3 4.2 2.7 lysosomal-associated membrane protein 3 194477 9.4 3.3 3.0 29.5 4.7 5.1 Yes Yes kinesin family member 20A 118830 3.3 2.5 8.6 serum deprivation response 156433 20.7 3.2 2.1 propionyl Coenzyme A carboxylase, alpha polypeptide 126484 3.6 3.2 2.1 acyl-CoA synthetase long-chain family member 5 180740 8.4 3.2 2.6 formin binding protein 1 127600 3.2 2.5 59.6 cyclin-dependent kinase inhibitor 1B (p27, Kip1) 121695 6.0 2.9 3.1 nudix (nucleoside diphosphate linked moiety X)-type 6 179802 2.1 3.4 3.1 DNA-damage-inducible transcript 4-like 199024 11.0 2.2 3.1 suppressor of cytokine signaling 2 108934 3.1 3.1 9.2 chromosome 20 open reading frame 35 220840 2.3 4.8 3.1 BTB (POZ) domain containing 3 119684 2.2 3.4 3.1 HRAS-like suppressor 3 144399 7.7 3.0 2.2 serine/threonine kinase 32B 178424 3.4 3.0 2.1 chromosome 6 open reading frame 85 182938 8.8 3.0 2.6 fibroblast growth factor 13 156333 11.6 3.0 2.2 protein tyrosine phosphatase, receptor N polypeptide 2 187260 4.1 3.0 2.7 zinc finger protein 435 101333 8.3 2.9 2.2 erythrocyte membrane protein band 4.1-like 2 124368 3.8 2.8 2.9 pellino homolog 1 (Drosophila) 113197 4.1 2.9 2.3 cAMP responsive element binding protein 3-like 4 203282 2.9 3.0 2.6 nuclear factor I/B 141557 10.8 2.9 2.6 F-box and leucine-rich repeat protein 7 178637 2.9 3.0 2.1 recombination activating gene 1 activating protein 1 210819 2.8 2.2 2.9 154 LEC gene signature (236) Fold change by AB1700 Fold change by qPCR Gene name AB Sample Sample Sample Sample Sample Sample LD- Core probe ID 1 2 3 1 2 3 MDA protein kinase C binding protein 1 180118 3.2 2.4 2.8 calsyntenin 3 198222 2.8 2.0 2.9 homeo box D1 149127 11.8 2.8 2.5 progestin and adipoQ receptor family member VIII 166596 13.1 2.8 2.1 t-complex 11 (mouse) like 2 193950 3.1 2.5 2.8 high-mobility group box 2 177404 2.0 2.8 8.2 glycine receptor, beta 112304 2.8 3.5 2.1 4.8 4.0 2.1 Yes contactin associated protein-like 3 153182 9.7 2.8 2.5 CNKSR family member 3 173337 5.8 2.3 2.8 F-box protein 46 162992 2.2 2.8 9.0 sprouty homolog 3 (Drosophila) 134280 2.8 2.6 3.0 3-hydroxybutyrate dehydrogenase (heart, mitochondrial) 122515 2.7 2.7 7.0 nudix (nucleoside diphosphate linked moiety X)-type 4 172746 4.7 2.7 2.4 Friend leukemia virus integration 1 128011 7.9 2.7 2.3 BMP2 inducible kinase 162982 4.3 2.3 2.7 phosphatidic acid phosphatase type 2A 207194 4.5 2.2 2.7 GTP cyclohydrolase I feedback regulator 215327 2.6 4.0 2.2 mutS homolog 5 (E. coli) 114236 2.6 2.5 3.0 Ras protein-specific guanine nucleotide-releasing factor 181938 2.6 2.6 2.2 2 transcription elongation factor B polypeptide 3B ( A2) 212385 2.6 2.1 11.2 gamma-aminobutyric acid (GABA) A receptor, epsilon 130287 2.5 2.6 2.6 ankylosis, progressive homolog (mouse) 172690 2.5 2.8 2.6 protein tyrosine phosphatase, non-receptor type 14 186519 2.8 2.5 2.5 cyclin B2 169571 2.5 2.1 8.3 semaphorin 3A 154374 2.5 2.5 5.8 cyclin-dependent kinase inhibitor 2C (p18, inhibits 112997 2.2 2.5 8.0 CDK4) MyoD family inhibitor domain containing 191935 2.5 2.5 3.1 Werner syndrome 188301 2.5 3.0 2.3 midline 2 206942 4.9 2.5 2.5 nuclear factor of activated T-cells, calcineurin- 198298 3.2 2.5 2.0 dependent 3 protein phosphatase 3 (calcineurin A alpha) 215875 4.3 2.4 2.2 adaptor-related protein complex 1, sigma 2 subunit 183970 2.4 2.4 2.4 SEC15-like 1 (S. cerevisiae) 147612 7.5 2.3 2.4 nucleoredoxin 160193 2.6 2.4 2.1 MAP kinase interacting serine/threonine kinase 2 147930 2.8 2.4 2.2 protein tyrosine phosphatase type IVA, member 3 221359 2.4 2.6 2.1 hydroxysteroid (17-beta) dehydrogenase 7 149125 2.2 3.1 2.3 TBC1 domain family, member 8 (with GRAM domain) 214636 8.1 2.3 2.2 citrate lyase beta like 124662 2.3 2.2 4.1 Rho GTPase activating protein 25 198348 14.4 2.3 2.3 forkhead box P1 203381 2.4 2.1 2.3 chromosome 14 open reading frame 94 187081 2.3 2.0 2.9 fat-like cadherin FATJ 131034 2.7 2.3 2.2 Fanconi anemia, complementation group C 205301 2.5 2.3 2.2 phosphorylase, glycogen; liver 150387 2.5 2.2 2.2 protocadherin gamma subfamily C, 5 227600 4.4 2.0 2.2 phosphatidylinositol transfer protein, cytoplasmic 1 117142 2.8 2.2 2.2 155 LEC gene signature (236) Fold change by AB1700 Fold change by qPCR Gene name AB Sample Sample Sample Sample Sample Sample LD- Core probe ID 1 2 3 1 2 3 MDA papilin, proteoglycan-like sulfated glycoprotein 138764 4.3 1.4 2.2 8.3 2.0 2.1 Yes Yes inhibitor of growth family, member 3 168063 2.0 2.1 2.1 2-hydroxyphytanoyl-CoA lyase 138137 2.1 2.1 2.3 extracellular link domain containing 1 195865 43.0 1.0 2.0 312.0 3.5 15.2 Yes EPH receptor B2 190778 2.2 1.4 1.4 4.8 1.6 0.7 Yes angiopoietin 2 193875 3.8 0.5 1.0 10.1 0.9 1.3 Yes desmoplakin 170090 1.5 0.4 0.5 2.5 3.5 3.3 Yes Yes BEC signature genes (sorted by median of AB1700 fold change) confirmed by LD-MDA BEC gene signature (342) Fold change by AB1700 Fold change by qPCR Gene name AB Sample Sample Sample Sample Sample Sample LD- Core probe ID 1 2 3 1 2 3 MDA plasminogen activator, urokinase 208672 165.6 6.0 126.5 91.6 6.7 164.8 Yes Yes proprotein convertase subtilisin/kexin type 1 213177 96.2 8.4 91.1 brain expressed, X-linked 1 137034 250.2 88.3 7.0 collagen, type I, alpha 2 105493 344.2 86.9 3.3 685.5 71.9 5.3 Yes glutamine-fructose-6-phosphate transaminase 2 113797 155.1 66.0 3.6 lipase, endothelial 200619 59.5 112.0 50.6 16.4 117.4 95.8 Yes Yes FAT tumor suppressor homolog 1 (Drosophila) 131558 54.8 54.6 8.7 6993.8 55.3 151.9 Yes Yes fibroblast activation protein, alpha 164725 131.1 47.4 46.7 16098.5 18.7 9098.3 Yes transgelin 172572 56.9 7.0 47.1 matrix metalloproteinase 1 (interstitial collagenase) 215808 45.6 9.5 85.9 36.3 11.7 73.3 Yes Yes ADAM with thrombospondin type 1 motif, 1 216353 477.8 45.4 4.4 140.1 64.1 3.7 Yes collagen, type VI, alpha 3 115643 884.5 44.2 2.8 interleukin 6 (interferon, beta 2) 163241 28.4 40.4 45.8 521.0 52.9 32.1 Yes Yes interleukin 1 receptor-like 1 131513 62.8 36.3 24.5 fms-related tyrosine kinase 1 219494 93.1 21.3 35.7 75.6 19.8 65.3 Yes Yes hypothetical protein BC012029 150646 71.0 9.0 34.9 interleukin 8 176899 34.1 9.0 33.2 26.8 7.6 68.7 Yes tumor necrosis factor (ligand) superfamily, member 15 192558 39.1 8.4 32.7 27.3 5.1 139.5 Yes Yes GLI pathogenesis-related 1 (glioma) 117689 118.7 32.7 9.8 collagen, type V, alpha 1 110570 68.3 5.7 31.0 36.5 5.3 83.6 Yes Yes cysteine-rich secretory protein LCCL domain 170538 331.3 29.5 5.4 containing 2 angiopoietin-like 4 181959 34.5 25.4 27.6 16.7 15.9 31.6 Yes Yes receptor tyrosine kinase-like orphan receptor 1 194924 51.8 8.2 27.0 brain abundant, membrane attached signal protein 1 198318 69.8 7.6 25.8 GLI-Kruppel family member GLI3 100093 24.0 2.7 31.3 latent transforming growth factor beta binding protein 2 116441 23.5 5.5 80.2 oxytocin receptor 200205 155.0 23.4 3.5 solute carrier family 22, member 17 140114 23.0 7.4 34.7 15.8 35.2 166.8 Yes Yes synaptotagmin-like 2 118410 115.0 22.5 17.6 chondroitin sulfate proteoglycan 2 (versican) 207524 117.5 21.9 4.8 4986.4 20.7 347.8 Yes glutaminyl-peptide cyclotransferase (glutaminyl 152127 25.4 4.0 20.6 cyclase) 156 BEC gene signature (342) Fold change by AB1700 Fold change by qPCR Gene name AB Sample Sample Sample Sample Sample Sample LD- Core probe ID 1 2 3 1 2 3 MDA glycoprotein (transmembrane) nmb 161212 236.8 20.3 4.1 11700.5 2503.8 42.8 Yes Yes high mobility group AT-hook 2 105728 20.3 20.2 3.0 brain-derived neurotrophic factor 215284 36.8 18.9 8.3 interferon, alpha-inducible protein 27 152567 18.9 7.3 170.0 lysyl oxidase-like 1 156579 102.7 8.2 17.7 transcription elongation factor A (SII)-like 7 130055 53.2 9.4 17.6 cadherin 2, type 1, N-cadherin (neuronal) 187321 163.7 5.2 17.5 24.5 3.4 14.7 Yes Yes dihydropyrimidinase-like 4 161807 17.2 4.7 22.2 phosphatidic acid phosphatase type 2 domain 190799 26.1 15.7 3.8 containing 1 stanniocalcin 2 184148 39.4 15.7 4.5 regulator of G-protein signalling 4 165955 93.6 15.5 10.9 squamous cell carcinoma antigen recognized by T cells 187160 15.7 4.6 15.5 2 chromosome 7 open reading frame 10 180432 87.0 15.5 13.3 GULP, engulfment adaptor PTB domain containing 1 137136 18.1 4.6 15.4 chromosome 6 open reading frame 105 163966 15.2 10.6 24.4 thrombospondin, type I, domain containing 2 155355 21.9 14.8 11.6 transient receptor potential cation channel, subfamily 158239 14.6 5.5 16.7 C1 epithelial membrane protein 3 152376 64.9 11.7 14.5 neuronal cell adhesion molecule 106462 14.4 7.8 29.6 11.5 7.9 26.7 Yes Yes leucine rich repeat containing 17 133553 14.1 4.0 48.6 nuclear receptor interacting protein 3 102896 19.2 9.0 13.8 ring finger protein 182 106266 31.3 5.6 13.7 neuregulin 1 223108 159.2 13.4 8.4 LY6/PLAUR domain containing 1 197493 7.2 12.3 14.1 lymphocyte cytosolic protein 1 (L-plastin) 175091 57.2 4.0 12.1 endothelial cell-specific molecule 1 174810 16.1 12.1 6.1 epithelial V-like antigen 1 129035 12.1 5.2 14.2 serine (or cysteine) proteinase inhibitor, clade E, 210342 15.7 3.5 12.0 member 2 phosphoglycerate dehydrogenase 210225 33.3 11.9 6.9 Fc receptor-like and mucin-like 2 173350 30.7 11.8 2.6 anthrax toxin receptor 1 112158 20.3 11.7 9.0 v-erb-b2 erythroblastic leukemia viral oncogene 105627 11.6 2.8 21.4 homolog 2 keratin 7 207298 11.5 9.3 21.3 IGF-II mRNA-binding protein 3 210073 12.5 6.8 11.4 inhibin, beta A (activin A, activin AB alpha 193064 50.6 5.8 11.2 44.0 4.6 9.4 Yes polypeptide) neurexin 3 139725 14.4 11.1 5.5 transcription factor EC 114012 11.1 6.2 45.3 transmembrane, prostate androgen induced RNA 158378 11.3 6.5 11.1 delta-notch-like EGF repeat-containing transmembrane 103310 35.7 10.1 10.9 1977.3 13.3 8699.8 Yes myozenin 2 188996 14.6 10.8 5.4 F-box protein 32 218680 10.7 4.0 16.2 CD44 antigen 133604 12.8 10.6 3.7 16.8 12.8 7.4 Yes sulfatase 1 169414 23.2 3.7 10.6 matrix metalloproteinase 10 (stromelysin 2) 170985 6.0 10.5 41.2 3.0 7.3 19.6 Yes heat shock 22kDa protein 8 165817 10.2 6.8 13.2 vascular endothelial growth factor C 170337 59.6 10.1 5.6 41.0 6.0 8.9 Yes Yes 157 BEC gene signature (342) Fold change by AB1700 Fold change by qPCR Gene name AB Sample Sample Sample Sample Sample Sample LD- Core probe ID 1 2 3 1 2 3 MDA shroom 207317 84.7 10.1 3.8 major facilitator superfamily domain containing 2 170831 25.2 10.1 2.5 Mst3 and SOK1-related kinase 112198 27.8 3.1 10.0 RAB23, member RAS oncogene family 122394 15.9 10.0 3.9 myosin, light polypeptide 9, regulatory 181719 12.4 4.0 9.9 intercellular adhesion molecule 1 (CD54) 109070 9.9 2.3 20.9 7.3 1.6 40.2 Yes nudix (nucleoside diphosphate linked moiety X)-type 125359 56.1 7.6 9.9 11 membrane metallo-endopeptidase (CALLA, CD10) 197353 117.8 4.7 9.8 7097.8 5.3 40.7 Yes solute carrier family 1, member 1 169469 10.6 9.7 2.5 cyclin-dependent kinase inhibitor 2B (p15, inhibits 204006 41.8 6.2 9.6 CDK4) spastic paraplegia 3A (autosomal dominant) 190761 10.4 2.1 9.3 multiple C2 domains, transmembrane 1 136033 4.4 22.8 9.3 lysyl oxidase-like 2 136648 34.1 3.6 9.2 synaptotagmin-like 3 171063 9.2 5.4 34.1 ecotropic viral integration site 1 201951 9.2 3.4 27.3 transforming growth factor, beta-induced, 68kDa 133906 52.8 9.1 8.0 spectrin domain with coiled-coils 1 103537 46.5 8.7 9.1 a disintegrin and metalloproteinase domain 23 177272 9.0 3.7 16.8 collagen, type VIII, alpha 1 219384 9.2 8.8 2.5 runt-related transcription factor 3 194489 15.5 8.7 2.2 fibulin 5 212132 8.5 4.4 19.1 family with sequence similarity 20, member C 199772 55.0 8.4 2.1 calmodulin binding transcription activator 1 103913 8.4 2.9 11.7 tropomyosin 2 (beta) 163441 28.4 8.4 5.1 leucine-rich repeat-containing G protein-coupled 155675 19.7 2.9 8.3 receptor 4 promethin 123271 8.3 4.2 22.5 sparc/osteonectin, cwcv and kazal-like domains 125730 8.8 4.3 8.2 proteoglycan nexilin (F actin binding protein) 139130 8.1 13.3 2.1 phospholipase A2, group IVA (cytosolic, calcium- 185568 4.1 18.6 7.9 dependent) prostaglandin-endoperoxide synthase 2 208388 3.3 7.9 8.9 interleukin 7 receptor 200834 54.3 7.7 2.9 37.0 10.7 9.1 Yes cadherin 11, type 2, OB-cadherin (osteoblast) 107867 21.2 7.7 2.1 17.4 12.1 3.6 Yes kelch-like 13 (Drosophila) 119847 4.7 8.4 7.7 zinc finger homeobox 1b 159875 46.6 7.6 5.2 guanylate binding protein 1, interferon-inducible, 145041 7.6 2.6 8.3 67kDa BCL2-associated athanogene 2 113353 17.6 7.5 2.6 amphoterin induced gene 2 154434 157.0 7.5 3.0 chromosome 14 open reading frame 37 211103 12.4 5.6 7.4 ring finger protein 150 213739 47.5 3.2 7.4 metallothionein 1F (functional) 144569 3.5 7.3 7.6 transforming growth factor, beta 3 158090 11.0 7.3 6.6 plasminogen activator, urokinase receptor 208060 14.2 7.2 6.1 serine (or cysteine) proteinase inhibitor, clade B, 183353 7.5 7.1 3.9 member 2 F-box and leucine-rich repeat protein 16 143089 7.1 2.0 11.6 UDP-N-acteylglucosamine pyrophosphorylase 1 120981 10.6 7.1 3.1 chromosome 9 open reading frame 150 194026 7.0 7.0 14.8 158 BEC gene signature (342) Fold change by AB1700 Fold change by qPCR Gene name AB Sample Sample Sample Sample Sample Sample LD- Core probe ID 1 2 3 1 2 3 MDA tumor necrosis factor receptor superfamily, member 6b 176555 7.5 3.0 7.0 asparagine synthetase 219578 13.3 4.0 7.0 piggyBac transposable element derived 3 218140 4.3 6.9 22.0 parathyroid hormone-like hormone 143454 8.8 6.8 2.4 insulin receptor substrate 1 183403 38.2 6.7 3.1 chromosome 6 open reading frame 115 116829 3.3 6.6 8.1 prostaglandin F receptor (FP) 103022 207.6 6.6 5.7 cysteine-rich secretory protein LCCL domain 113861 15.7 5.1 6.6 containing 1 programmed cell death 1 ligand 2 213263 6.4 15.1 3.9 dual specificity phosphatase 23 140616 12.4 6.4 5.1 ubiquitin carboxyl-terminal esterase L1 192628 6.4 2.5 13.2 tumor necrosis factor (ligand) superfamily, member 9 190964 7.0 6.2 2.7 chromosome 6 open reading frame 188 223249 6.7 6.2 5.4 ATPase, Na+/K+ transporting, beta 1 polypeptide 199586 6.2 7.6 3.9 myosin IE 142916 6.2 2.7 10.2 B-cell CLL/lymphoma 6 (zinc finger protein 51) 151724 7.9 4.9 6.0 follistatin-like 1 140364 9.4 2.4 6.0 vitamin D (1,25- dihydroxyvitamin D3) receptor 135316 44.7 5.9 2.5 sushi-repeat-containing protein, X-linked 2 140054 5.9 3.2 41.8 chondroitin sulfate synthase 3 223255 7.7 3.2 5.9 SH3 domain containing ring finger 2 131603 6.6 5.9 2.8 junctional adhesion molecule 3 100510 5.9 3.6 14.7 2.7 2.8 15.6 Yes Yes RGM domain family, member B 223528 12.8 2.7 5.9 discoidin, CUB and LCCL domain containing 2 114133 6.8 4.5 5.9 metallothionein 1B (functional) 174119 4.0 8.3 5.8 proteoglycan 1, secretory granule 133333 5.8 5.4 8.9 GATA binding protein 6 186512 11.2 5.8 4.9 synaptogyrin 1 223759 10.2 2.8 5.7 elongation of very long chain fatty acids-like 4 199358 10.8 5.5 5.6 interleukin 32 143239 5.0 5.6 35.0 N-acetyltransferase 2 (arylamine N-acetyltransferase) 105812 5.6 4.6 5.8 procollagen-proline, 2-oxoglutarate 4-dioxygenase 182482 6.6 2.7 5.5 metallothionein 1X 226343 4.1 8.2 5.5 nuclear transport factor 2 212373 4.4 7.7 5.5 palladin, cytoskeletal associated protein 130541 11.6 2.2 5.4 chemokine (C-X-C motif) ligand 1 149192 5.4 4.9 5.9 metallothionein 2A (functional) 204773 4.0 8.5 5.4 angiotensin II receptor, type 1 105197 39.6 5.3 3.2 regulator of G-protein signalling 5 192935 4.3 9.0 5.3 regulator of G-protein signalling 10 128800 10.5 5.2 4.9 ankyrin repeat domain 42 161072 9.8 2.6 5.2 Rho GDP dissociation inhibitor (GDI) beta 143589 2.3 5.2 5.5 regulator of G-protein signalling 20 138203 5.2 5.5 3.0 myeloid/lymphoid or mixed-lineage leukemia 151049 5.2 2.2 6.3 dihydropyrimidine dehydrogenase 112355 3.7 5.4 5.1 bone morphogenetic protein 8b (osteogenic protein 2) 112045 5.3 3.1 5.1 collagen, type VI, alpha 1 215580 63.7 5.0 4.0 39.4 7.5 14.3 Yes 159 BEC gene signature (342) Fold change by AB1700 Fold change by qPCR Gene name AB Sample Sample Sample Sample Sample Sample LD- Core probe ID 1 2 3 1 2 3 MDA pleckstrin homology-like domain, family A, member 2 185855 11.7 5.0 4.1 potassium large conductance calcium-activated channel 213246 5.0 5.2 3.4 (Mb1) pregnancy specific beta-1-glycoprotein 7 139074 2.8 5.1 4.8 lymphocyte antigen 6 complex, locus K 100594 17.2 3.6 4.8 DAZ interacting protein 1 115295 8.8 4.8 2.6 metallothionein IV 223241 2.6 4.8 5.1 basic helix-loop-helix domain containing, class B, 2 197538 8.9 3.2 4.8 matrix metalloproteinase 2 (gelatinase A) 146058 4.7 2.1 17.4 collagen, type XXVII, alpha 1 224496 6.8 3.1 4.7 solute carrier family 7, member 14 186645 4.6 6.7 3.9 SMAD specific E3 ubiquitin protein ligase 2 126729 5.0 4.6 2.8 metallothionein 1E (functional) 223856 3.1 6.6 4.6 ADP-ribosylation factor 7 210546 5.4 4.5 3.2 ATPase, Class I, type 8B, member 1 106683 4.5 2.3 6.7 interferon induced transmembrane protein 5 163015 2.9 4.7 4.5 chromosome X open reading frame 53 150469 9.7 2.2 4.5 pregnancy specific beta-1-glycoprotein 5 101520 4.4 9.2 2.7 solute carrier family 19 (thiamine transporter), member 119822 13.3 3.0 4.4 2 ATP-binding cassette, sub-family B (MDR/TAP), 182279 4.4 2.4 13.7 member 1 tumor necrosis factor receptor superfamily, member 137897 13.1 4.3 3.4 12A Notch homolog 2 (Drosophila) N-terminal like 128130 4.3 2.1 6.6 low density lipoprotein-related protein 12 114806 10.0 2.6 4.3 retinoic acid receptor, beta 109692 19.8 4.3 3.3 leucine rich repeat containing 16 141737 7.8 4.3 3.7 ankyrin repeat and SOCS box-containing 9 109802 2.8 4.3 14.8 integrin, beta 3 (platelet glycoprotein IIIa, antigen 192782 5.1 4.0 4.2 3.1 2.8 6.1 Yes CD61) serine (or cysteine) proteinase inhibitor, clade B 226878 11.7 4.2 2.5 member 8 cytoskeleton-associated protein 4 134467 4.2 2.6 4.5 natriuretic peptide precursor C 146370 21.3 4.2 4.1 tripartite motif-containing 61 229268 5.6 4.2 2.8 quinolinate phosphoribosyltransferase 161752 46.4 4.2 2.4 G protein-coupled receptor 8 169779 2.2 4.2 6.7 immunoglobulin superfamily, member 4B 149440 13.6 2.7 4.1 caldesmon 1 185731 4.7 3.3 4.1 activated leukocyte cell adhesion molecule 115975 6.5 4.1 2.4 PFTAIRE protein kinase 1 129209 3.3 4.1 4.4 TBC1 domain family, member 2 205982 6.9 4.1 2.6 selectin P 114371 4.0 3.7 14.3 guanylate binding protein 3 221503 5.0 2.5 4.0 secreted protein, acidic, cysteine-rich (osteonectin) 123437 4.0 2.4 9.2 cysteine and glycine-rich protein 1 203065 5.4 2.4 4.0 leukotriene B4 12-hydroxydehydrogenase 210882 4.0 2.7 5.1 microtubule-associated protein 1A 130759 32.7 3.9 2.9 carboxypeptidase A3 (mast cell) 100989 3.9 4.3 2.3 autism susceptibility candidate 2 156350 3.9 2.6 4.3 laminin, gamma 2 201627 3.9 5.8 3.4 160 BEC gene signature (342) Fold change by AB1700 Fold change by qPCR Gene name AB Sample Sample Sample Sample Sample Sample LD- Core probe ID 1 2 3 1 2 3 MDA family with sequence similarity 91, member A1 103214 2.6 4.9 3.9 BIA2 216429 3.8 8.8 3.2 zinc finger protein 568 190182 4.1 3.2 3.8 dual-specificity tyrosine-(Y)-phosphorylation kinase 2 130435 7.1 2.3 3.8 pleckstrin homology-like domain, family B, member 1 120125 3.8 2.4 3.9 metallothionein 1J 227956 3.6 11.6 3.7 2,3-bisphosphoglycerate mutase 178503 5.8 3.6 3.7 regulator of G-protein signalling 3 221266 3.7 2.0 25.5 armadillo repeat containing 9 141649 3.7 2.6 4.6 echinoderm microtubule associated protein like 1 119690 2.6 3.7 5.2 chromosome 6 open reading frame 168 177450 3.7 3.0 7.8 S100 calcium binding protein A3 199821 2.7 4.8 3.7 adenylate cyclase 7 184103 7.4 2.0 3.7 PDZ and LIM domain 7 (enigma) 121548 5.4 2.5 3.6 tropomyosin 1 (alpha) 225735 7.0 3.6 2.8 interleukin 4 receptor 214374 3.6 2.4 17.8 leukocyte receptor cluster (LRC) member 4 212382 3.6 2.0 6.4 plasminogen activator, tissue 152725 22.2 3.6 3.6 piggyBac transposable element derived 5 133539 5.0 2.4 3.6 exostoses (multiple) 1 190050 5.4 3.6 2.2 ankyrin repeat domain 1 (cardiac muscle) 209661 3.6 6.1 2.7 pregnancy specific beta-1-glycoprotein 9 130276 15.0 2.4 3.6 UL16 binding protein 2 221382 5.9 3.5 3.0 S100 calcium binding protein A11 (calgizzarin) 145550 3.5 2.7 3.7 glypican 2 (cerebroglycan) 184349 3.5 3.1 14.9 protein kinase, cAMP-dependent, catalytic, beta 198878 2.2 3.5 4.9 Rho GTPase activating protein 24 111787 3.5 2.6 6.1 testis expressed gene 9 102862 11.4 3.1 3.5 proline-serine-threonine phosphatase interacting protein 141843 3.5 3.5 3.6 2 integrin, alpha V 117958 3.5 2.6 6.1 2.0 2.6 5.0 Yes fibronectin 1 136386 21.2 3.5 2.5 SPOC domain containing 1 162279 4.9 3.5 3.5 protein tyrosine phosphatase, receptor type, G 171586 3.5 2.6 3.6 fibronectin type III and ankyrin repeat domains 1 173923 3.5 2.5 5.6 regulating synaptic membrane exocytosis 2 216707 5.2 3.4 2.8 C2 and WW domain containing E3 ubiquitin protein 103552 3.2 3.4 3.7 ligase 2 La ribonucleoprotein domain family, member 6 150810 8.2 2.4 3.4 ribonuclease, RNase A family, 1 (pancreatic) 139103 3.4 3.3 197.0 opioid growth factor receptor-like 1 117428 3.4 2.8 3.6 adrenergic, beta, receptor kinase 2 213717 4.0 2.2 3.4 histone 1, H2bk 190073 3.3 2.5 3.7 meningioma (disrupted in balanced translocation) 1 203105 2.9 3.4 3.3 poliovirus receptor 149750 6.4 3.3 2.2 sine oculis homeobox homolog 1 (Drosophila) 113710 5.1 3.3 2.6 chemokine (C-X-C motif) ligand 6 192528 22.7 3.3 2.1 cyclin-dependent kinase inhibitor 2A 127523 8.6 3.3 2.5 homeodomain interacting protein kinase 2 233986 3.7 2.1 3.3 161 BEC gene signature (342) Fold change by AB1700 Fold change by qPCR Gene name AB Sample Sample Sample Sample Sample Sample LD- Core probe ID 1 2 3 1 2 3 MDA collagen, type V, alpha 3 181976 3.3 2.1 7.8 Notch homolog 2 (Drosophila) 142906 10.7 2.8 3.3 ectodermal-neural cortex (with BTB-like domain) 172733 7.5 3.3 3.0 cyclin-dependent kinase 6 235779 5.6 2.5 3.3 enabled homolog (Drosophila) 219949 3.1 3.3 5.6 retinoic acid early transcript 1G 177046 5.3 3.2 3.2 tubulin, beta 2 194068 3.9 3.2 2.4 spastic ataxia of Charlevoix-Saguenay (sacsin) 211211 4.8 3.2 2.9 colony stimulating factor 2 (granulocyte-macrophage) 155234 3.2 2.4 8.1 phosphatidylinositol transfer protein, membrane- 128232 3.5 2.3 3.2 associated 1 solute carrier family 7, member 11 209308 3.2 4.0 2.9 ventricular zone expressed PH domain homolog 1 178585 3.1 2.9 6.2 S100 calcium binding protein A11 pseudogene 187317 3.1 2.7 6.2 homeo box B2 167151 4.8 3.1 3.1 glucan (1,4-alpha-), branching enzyme 1 234875 2.3 3.6 3.1 discoidin, CUB and LCCL domain containing 1 211479 5.5 3.0 2.5 ras homolog gene family, member Q 142100 3.6 2.9 3.0 activating transcription factor 3 185687 3.3 2.4 3.0 zinc finger protein 528 121921 2.2 3.0 3.6 G protein-coupled receptor 126 134617 3.0 3.1 2.3 AIF-like mitochondrion-associated inducer of death 129590 3.0 2.0 3.9 collagen triple helix repeat containing 1 191084 3.0 2.2 11.0 protein tyrosine phosphatase-like A domain containing 205747 2.1 3.0 5.4 2 Rho-related BTB domain containing 3 103105 2.9 2.8 5.1 IBR domain containing 2 234266 4.3 2.9 2.9 ATPase family homolog up-regulated in senescence 222770 2.9 3.1 2.1 cells protocadherin 10 116092 12.1 2.8 2.7 protein kinase, AMP-activated, beta 2 non-catalytic 180648 2.4 2.8 7.5 subunit apolipoprotein B mRNA editing enzyme 132926 2.1 2.8 3.1 ecotropic viral integration site 5 124405 2.8 2.8 3.3 glycosyltransferase-like domain containing 1 215619 2.8 2.0 4.3 pregnancy specific beta-1-glycoprotein 1 211648 2.6 5.5 2.8 diacylglycerol kinase, gamma 90kDa 211883 2.0 2.8 6.0 protocadherin beta 17 pseudogene 218372 3.4 2.3 2.7 paired-like homeodomain transcription factor 2 167147 16.0 2.7 2.7 chromosome 6 open reading frame 128 223051 2.3 2.7 11.4 phosphatidylserine receptor 109646 2.5 2.7 4.6 pre-B-cell leukemia transcription factor 3 168129 3.2 2.7 2.6 protocadherin beta 5 169402 2.4 2.7 12.5 mitogen-activated protein kinase kinase kinase 13 143560 2.6 2.6 6.2 a disintegrin and metalloproteinase domain 9 183965 3.1 2.3 2.6 IKK interacting protein 223807 2.9 2.6 2.3 cathepsin C 190056 2.6 2.6 3.8 protein kinase C, epsilon 214787 2.6 2.6 7.2 chemokine (C-C motif) receptor-like 2 112106 2.5 2.6 11.6 membrane protein, palmitoylated 2 (MAGUK p55 152735 10.5 2.6 2.5 subfamily member 2) 162 BEC gene signature (342) Fold change by AB1700 Fold change by qPCR Gene name AB Sample Sample Sample Sample Sample Sample LD- Core probe ID 1 2 3 1 2 3 MDA phosphoprotein enriched in astrocytes 15 220353 3.6 2.1 2.6 transcription elongation factor A (SII)-like 5 157065 2.5 2.6 6.5 transient receptor potential cation channel, subfamily V 207656 2.5 2.1 7.4 2 phosphoglucomutase 3 205937 2.7 2.5 2.1 chaperonin containing TCP1, subunit 6B (zeta 2) 130061 2.5 2.1 7.6 trophoblast-derived noncoding RNA 232041 2.5 2.4 3.3 signal-induced proliferation-associated 1 like 2 169848 2.3 2.5 4.6 muscle RAS oncogene homolog 101584 3.1 2.4 2.5 ecotropic viral integration site 2B 139158 8.8 2.3 2.4 insulin-like growth factor binding protein 6 150353 9.5 2.4 2.0 serine-arginine repressor protein (35 kDa) 106827 2.4 2.0 2.5 coagulation factor II (thrombin) receptor 158262 2.4 2.2 6.4 fibrillin 1 (Marfan syndrome) 147111 10.0 2.4 2.1 CD109 antigen (Gov platelet alloantigens) 157795 2.4 2.4 3.9 four and a half LIM domains 2 202137 4.1 2.4 2.4 TRAF family member-associated NFKB activator 109879 2.3 2.0 4.9 anthrax toxin receptor 2 191543 2.3 2.7 2.1 membrane protein, palmitoylated 4 150693 2.3 2.2 3.8 PDZ and LIM domain 5 103100 2.3 3.0 2.3 fem-1 homolog c (C.elegans) 127238 4.9 2.3 2.3 nicotinamide N-methyltransferase 125400 2.6 2.3 2.1 protein tyrosine phosphatase domain containing 1 128960 2.3 2.1 8.2 golgi membrane protein SB140 151061 2.3 2.2 2.5 congenital dyserythropoietic anemia, type I 114892 2.2 2.3 2.3 DnaJ (Hsp40) homolog, subfamily B, member 5 152148 2.1 2.3 3.8 solute carrier family 38, member 6 142416 2.3 2.3 2.3 C1q domain containing 1 106320 2.2 2.3 2.4 palmdelphin 114289 2.1 2.3 2.8 related RAS viral (r-ras) oncogene homolog 180365 2.1 2.3 6.4 SNF1-like kinase 171526 19.1 2.2 2.2 solute carrier family 35, member D1 135060 2.1 2.6 2.1 solute carrier family 16, member 3 129361 6.6 2.1 2.1 retinoic acid early transcript 1K pseudogene 190201 2.2 2.0 2.1 163 Appendix Figure 1 Prediction Relevance Ranking analysis. A novel method denoted “Prediction Relevance Ranking” analysis employs multiple linear regression analysis and it enumerates all possible models and investigates the whole model space to produce a ranking of variables (genes) based on their predictive power. The prediction relevance for lymphatic vessel area (LVA; blue bars) revealed INHBA, DPP4, IL7R, TEK and ADAMTS1 as top 5 most frequent variables in a set of all possible regression models. The prediction relevance for blood vessel area (BVA; red bars) revealed FLT1, CSPG2, ITGB3, FGF12 and IL8 as top 5 most frequent variables in a set of all possible significant regression models. LVA BVA o 0 0 0 0 0 000000 o 0 0 0 0 o 0 0 0 0 o 0 0 0 0 o 0 0 0 0 T""" C\J Cf) '<:t L() T""" C\J Cf) '<:t L() Count Count 164 Appendix Table 2 Top 400 differentially modulated genes by VEGF-A in LEC using Multivariate Bayesian ranking analysis LEC stimulated with VEGF-A Log2 Ratio MB Ranking Name Probe ID 1 hr 4 hr. 8 hr. 24 hr. 1 early growth response 3 124744 7.306 0.238 0.308 0.463 2 early growth response 2 (Krox-20 homolog, Drosophila) 101929 6.522 -0.414 -0.206 -0.285 3 early growth response 1 147353 4.671 0.879 1.207 0.623 4 coagulation factor III (thromboplastin, tissue factor) 204787 5.860 3.093 1.424 1.559 5 N/A 209213 2.357 -1.128 -1.041 -1.699 6 activating transcription factor 3 185687 4.483 0.985 0.728 0.931 7 nuclear receptor subfamily 4, group A, member 2 123450 5.775 -0.154 0.051 0.558 8 nuclear receptor subfamily 4, group A, member 1 216600 6.072 1.846 0.123 0.653 9 v-fos FBJ murine osteosarcoma viral oncogene homolog 205128 4.676 1.846 1.819 1.243 10 N/A 183303 2.878 1.011 0.477 0.115 11 Kruppel-like factor 10 104402 2.859 0.929 0.491 -0.001 12 Kruppel-like factor 10 125028 2.834 0.845 0.630 0.193 13 coagulation factor III (thromboplastin, tissue factor) 146916 5.360 2.604 1.229 1.460 14 N/A 122171 2.095 0.146 -0.173 -0.744 15 jumonji domain containing 3 124424 2.168 -0.445 -0.869 -0.484 16 cytochrome P450, family 1, subfamily A, polypeptide 1 135086 -0.049 -1.869 -2.830 -2.051 17 prostaglandin-endoperoxide synthase 2 (prostaglandin G/H synthase 208388 4.575 -0.148 0.357 -0.137 and cyclooxygenase) 18 family with sequence similarity 13, member C1 205995 0.319 -1.749 -1.979 -0.823 19 dual specificity phosphatase 5 121612 3.317 2.588 1.487 0.938 20 N/A 178581 2.694 1.288 0.516 0.599 21 zinc finger protein 36, C3H type, homolog (mouse) 179827 3.494 1.279 1.444 1.151 22 guanylate cyclase 1, soluble, alpha 3 170165 0.680 0.940 -1.154 -1.715 23 SHC (Src homology 2 domain containing) family, member 4 183641 0.344 2.615 1.045 0.400 24 chromosome 11 open reading frame 17|NUAK family, SNF1-like 157942 2.637 0.647 0.354 0.482 kinase, 2 25 hairy and enhancer of split 1, (Drosophila) 176983 1.842 -0.335 -0.006 -0.228 26 kinesin family member 20A 118830 -1.084 -1.623 -1.397 0.728 27 solute carrier family 2 (facilitated glucose transporter), member 12 175805 -0.120 -1.846 -2.481 -1.424 28 Down syndrome critical region gene 1 124953 4.418 1.431 0.768 0.589 29 hairy/enhancer-of-split related with YRPW motif 1 121998 1.373 -1.139 -0.539 -1.108 30 kinesin family member 4A 115354 -0.777 -1.145 -0.719 1.225 31 ribosomal protein S6 kinase, 90kDa, polypeptide 5 152363 0.643 -0.738 -2.405 -0.106 32 carbonic anhydrase IV 196942 0.055 -0.351 -1.447 -3.247 33 topoisomerase (DNA) II alpha 170kDa 135302 -0.165 -0.395 -0.739 1.851 34 cell division cycle associated 1 147806 -0.539 -0.869 -0.625 1.457 35 SMAD, mothers against DPP homolog 7 (Drosophila) 206947 0.715 -1.260 -1.112 -0.860 36 centromere protein A, 17kDa 128411 -0.901 -0.730 -0.705 1.393 37 discs, large homolog 7 (Drosophila) 194498 -0.342 -0.630 -0.911 1.458 38 stanniocalcin 1 119453 4.020 2.647 -0.120 0.651 39 dual specificity phosphatase 1 182417 2.918 2.217 1.441 1.207 40 N/A 138125 1.170 1.633 0.265 -0.877 165 LEC stimulated with VEGF-A Log2 Ratio MB Ranking Name Probe ID 1 hr 4 hr. 8 hr. 24 hr. 41 centromere protein F, 350/400ka (mitosin) 183726 0.073 -0.272 -0.571 1.796 42 DEP domain containing 1B 206865 -0.544 -0.975 -1.279 1.065 43 aurora kinase B 203163 -0.610 -0.709 -0.636 1.537 44 protein regulator of cytokinesis 1 180626 -0.726 -0.751 -0.563 1.512 45 centrosomal protein 55kDa 198728 -0.492 -0.734 -0.553 1.480 46 solute carrier family 4, sodium bicarbonate cotransporter, member 7 163733 1.259 2.638 1.880 1.292 47 nuclear factor, interleukin 3 regulated 203863 1.709 0.030 0.108 -0.221 48 N/A 190308 -0.252 -0.685 -0.222 1.737 49 TTK protein kinase 107112 -0.521 -0.850 -0.278 1.577 50 NIMA (never in mitosis gene a)-related kinase 2 115004 -0.918 -1.137 -1.067 0.987 51 kinesin family member 2C 212531 -0.376 -0.587 -0.535 1.644 52 N/A 111700 -0.199 -0.568 0.036 1.829 53 apolipoprotein B mRNA editing enzyme, catalytic polypeptide-like 3B 138620 -0.916 -1.347 -0.883 1.068 54 baculoviral IAP repeat-containing 3 154663 2.395 -0.108 -0.044 0.036 55 SNF1-like kinase 171526 2.704 0.921 0.369 0.334 56 nucleolar and spindle associated protein 1 128435 -0.424 -0.631 -0.721 1.657 57 FBJ murine osteosarcoma viral oncogene homolog B 105390 5.350 0.229 0.031 -0.222 58 transforming, acidic coiled-coil containing protein 3 165983 -0.825 -1.252 -0.506 0.944 59 chemokine orphan receptor 1 139192 1.941 1.132 0.255 -0.142 60 mitogen-activated protein kinase kinase kinase 8 106192 2.404 0.220 -0.228 0.575 61 trophinin associated protein (tastin) 137875 -1.182 -1.208 -0.882 0.800 62 PDZ binding kinase 169723 -1.125 -1.005 -0.883 1.315 63 cyclin-dependent kinase inhibitor 2C (p18, inhibits CDK4) 112997 -0.573 -1.672 -1.590 0.539 64 N/A 125681 0.980 1.524 0.197 -0.832 65 pyruvate dehydrogenase kinase, isozyme 4 101060 -0.369 -1.822 -2.025 -1.423 66 asp (abnormal spindle)-like, microcephaly associated (Drosophila) 181685 0.162 -0.066 -0.278 2.022 67 N/A 133024 -0.642 -2.246 -2.748 -3.315 68 cyclin A2 110863 -0.567 -0.681 -0.593 1.370 69 sprouty homolog 2 (Drosophila) 207231 1.994 0.659 0.868 0.388 70 baculoviral IAP repeat-containing 5 (survivin) 104062 -0.680 -0.912 -0.807 1.171 71 chemokine (C-X-C motif) receptor 4 191821 -0.397 0.934 1.974 2.182 72 fms-related tyrosine kinase 1 219494 0.396 2.332 1.457 1.320 73 proline rich 11 123490 -0.514 -0.909 -0.813 1.123 74 polo-like kinase 1 (Drosophila) 197341 -0.897 -0.863 -0.985 0.921 75 N/A 198371 -0.997 -1.140 -1.770 0.442 76 spindle pole body component 25 homolog (S. cerevisiae) 130624 -0.592 -1.014 -0.793 1.178 77 SHC SH2-domain binding protein 1 106848 -0.572 -0.523 0.134 1.726 78 chromosome 18 open reading frame 24 205780 -0.714 -0.946 -0.345 1.225 79 prickle-like 2 (Drosophila) 134212 0.046 -1.868 -1.507 -0.900 80 anillin, actin binding protein (scraps homolog, Drosophila) 169499 -0.562 -0.488 -0.450 1.631 81 v-maf musculoaponeurotic fibrosarcoma oncogene homolog B (avian) 226336 1.215 -0.989 -0.899 -0.973 82 N/A 200967 2.667 0.366 -0.471 -0.649 83 MAX dimerization protein 3 131913 -0.523 -1.359 -1.663 0.399 84 high mobility group AT-hook 2 105728 -0.182 2.062 1.305 0.270 85 N/A 715431 0.288 1.100 1.165 -0.858 86 family with sequence similarity 64, member A 201158 -0.802 -0.532 -1.105 0.913 87 heparin-binding EGF-like growth factor 139874 2.354 1.481 0.874 1.046 166 LEC stimulated with VEGF-A Log2 Ratio MB Ranking Name Probe ID 1 hr 4 hr. 8 hr. 24 hr. 88 barren homolog 1 (Drosophila) 177741 -0.581 -0.687 -0.161 1.426 89 phosphoinositide-3-kinase, regulatory subunit 1 (p85 alpha) 180157 1.687 0.505 0.063 -0.259 90 N/A 170545 0.957 1.485 0.132 -0.930 91 kinetochore associated 2 107406 -0.810 -1.111 -0.986 1.101 92 v-myb myeloblastosis viral oncogene homolog (avian)-like 1 207803 -0.522 -1.318 0.036 0.737 93 BUB1 budding uninhibited by benzimidazoles 1 homolog (yeast) 157194 -1.122 -0.989 -0.638 0.826 94 kinesin family member 23 160577 -0.457 -0.538 -0.242 1.513 95 phosphoglycerate dehydrogenase 210225 -0.403 -0.621 -0.224 1.452 96 hyaluronan-mediated motility receptor (RHAMM) 216917 -0.450 -0.559 -0.537 1.550 97 N/A 215658 1.332 2.110 1.582 1.585 98 thymidine kinase 1, soluble 105119 -0.759 -0.809 -0.475 1.377 99 kinesin family member 18A 177455 -0.158 -0.801 -0.649 1.446 100 prickle-like 1 (Drosophila) 143140 1.064 -0.369 -0.876 -0.308 101 chromosome 15 open reading frame 42 196613 -0.844 -0.068 0.454 1.578 102 Opa interacting protein 5 211010 -0.539 -0.800 -0.980 1.090 103 B-cell CLL/lymphoma 6 (zinc finger protein 51) 151724 2.292 0.323 0.285 0.708 104 cyclin B2 169571 -0.727 -1.086 -1.148 0.980 105 N/A 179136 1.633 -0.311 -0.205 0.199 106 adrenergic, beta-2-, receptor, surface 127856 1.102 0.315 -0.929 -0.437 107 glyoxalase domain containing 1 137815 -0.591 1.175 0.669 -0.469 108 N/A 159898 -0.868 -0.645 -0.773 1.058 109 endothelial cell-specific molecule 1 174810 0.542 2.492 2.233 3.508 110 kinesin family member 11 199107 -0.325 -0.596 -0.446 1.497 111 phosphoserine aminotransferase 1 151268 -0.433 0.255 1.460 1.317 112 kinesin family member C1 141343 -0.547 -0.719 -0.524 1.179 113 leukemia inhibitory factor (cholinergic differentiation factor) 117096 2.348 0.370 0.269 0.026 114 ubiquitin-conjugating enzyme E2C 143651 -0.417 -0.478 -0.491 1.299 115 baculoviral IAP repeat-containing 5 (survivin) 227666 -0.696 -0.628 -0.772 1.018 116 N/A 134295 0.213 -1.245 -1.175 -0.003 117 basic helix-loop-helix domain containing, class B, 2 197538 3.844 1.508 1.612 1.404 118 MLF1 interacting protein 147135 -0.525 -0.699 0.094 1.405 119 solute carrier family 38, member 4 217080 0.455 -0.846 -1.692 -0.761 120 proline/serine-rich coiled-coil 1 181161 -0.794 -1.379 -1.233 0.532 121 Rac GTPase activating protein 1 135746 -0.553 -0.694 -0.499 1.086 122 inositol 1,4,5-trisphosphate 3-kinase A 216718 0.277 1.824 1.149 -0.143 123 cyclin-dependent kinase inhibitor 3 (CDK2-associated dual specificity 191305 -0.512 -0.579 -0.702 1.212 phosphatase) 124 MAP6 domain containing 1 180319 -0.055 1.554 0.545 0.316 125 transmembrane protein 100 216519 -0.288 -1.704 -1.210 -1.785 126 asp (abnormal spindle)-like, microcephaly associated (Drosophila) 205649 0.013 -0.681 -0.595 1.570 127 tumor necrosis factor, alpha-induced protein 8 128843 1.698 -0.128 -0.095 0.149 128 N/A 217029 2.200 0.977 0.786 0.738 129 N/A 128495 -0.589 -0.539 -0.329 1.395 130 amphiregulin (schwannoma-derived growth factor) 123143 1.383 1.936 0.394 0.351 131 serine/threonine kinase 6 pseudogene 683679 -0.650 -0.693 -0.475 1.028 132 cadherin 10, type 2 (T2-cadherin) 198370 0.309 -1.144 -1.206 -0.838 133 cell division cycle 2, G1 to S and G2 to M 123400 -0.457 -0.656 -0.421 1.440 134 antigen identified by monoclonal antibody Ki-67 137656 -0.521 -0.679 -0.476 1.495 167 LEC stimulated with VEGF-A Log2 Ratio MB Ranking Name Probe ID 1 hr 4 hr. 8 hr. 24 hr. 135 high-mobility group box 2 216889 -0.690 -1.169 -1.097 0.652 136 protein tyrosine phosphatase, receptor type, E 541414 2.225 0.800 1.020 0.127 137 ribonucleotide reductase M2 polypeptide 134286 -0.576 -0.569 -0.363 1.698 138 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3 198714 1.992 1.866 1.293 0.737 139 ephrin-B2 162498 0.755 -0.516 -1.167 -0.835 140 myeloid cell leukemia sequence 1 (BCL2-related) 147139 1.818 0.453 0.337 0.105 141 v-rel reticuloendotheliosis viral oncogene homolog (avian) 109539 0.959 -0.687 -0.419 0.064 142 mal, T-cell differentiation protein-like 201519 -0.313 1.677 0.776 -0.142 143 N/A 163671 -0.172 1.423 1.141 0.763 144 Rho GTPase activating protein 11A 117485 -0.073 0.011 0.318 1.696 145 pituitary tumor-transforming 3 147452 -0.575 -0.708 -0.912 0.854 146 polymerase (DNA directed), theta 112081 0.167 -0.429 0.290 1.452 147 midnolin 159555 2.462 0.864 0.542 0.391 148 potassium channel, subfamily K, member 6 206064 -0.536 1.242 0.908 0.487 149 chromosome 18 open reading frame 1 113714 0.132 -1.139 -1.452 -0.862 150 interleukin 1, beta 130322 2.151 0.852 0.563 0.950 151 cancer susceptibility candidate 5 105014 -0.450 -0.238 -0.317 1.485 152 phosphoserine aminotransferase 1 221068 -0.516 0.066 1.354 1.094 153 tumor necrosis factor, alpha-induced protein 3 168741 2.476 0.210 -0.071 0.099 154 ATPase family, AAA domain containing 3B 224173 -0.119 1.243 1.364 0.606 155 centromere protein A, 17kDa 106198 -0.873 -1.110 -0.739 1.096 156 sperm associated antigen 5 185888 -0.713 -0.663 -0.760 0.878 157 nucleoside phosphorylase 147282 1.265 1.971 1.564 1.491 158 pituitary tumor-transforming 1 204044 -0.603 -0.636 -0.787 0.935 159 gamma-aminobutyric acid (GABA) A receptor, epsilon 130287 1.430 0.864 0.351 -0.280 160 ADAM metallopeptidase with thrombospondin type 1 motif, 9 160876 2.459 0.884 0.422 -0.495 161 cyclin B1 216120 -0.852 -0.688 -0.560 0.812 162 nuclear receptor subfamily 1, group D, member 2 227856 0.585 -0.375 -1.385 -0.080 163 N/A 138397 -0.489 -0.303 0.355 1.295 164 MAD2 mitotic arrest deficient-like 1 (yeast) 138685 -0.392 -0.272 0.399 1.355 165 N/A 192975 -0.959 -1.012 -0.260 1.218 166 TPX2, microtubule-associated, homolog (Xenopus laevis) 189094 -0.848 -0.906 -0.409 1.022 167 DEP domain containing 1 205423 -0.278 -0.660 -0.356 1.362 168 v-maf musculoaponeurotic fibrosarcoma oncogene homolog (avian) 186589 0.095 -1.340 -1.613 -1.197 169 CDC45 cell division cycle 45-like (S. cerevisiae) 208272 -0.238 -0.308 0.532 1.740 170 CDC20 cell division cycle 20 homolog (S. cerevisiae) 161150 -0.771 -0.855 -0.725 0.734 171 sphingosine kinase 1 199219 -0.146 1.291 0.300 -0.128 172 regulator of G-protein signalling 16 173025 1.972 0.545 0.667 0.140 173 serine/threonine kinase 6 157917 -0.561 -0.644 -0.506 1.008 174 spermatogenesis associated 12 229375 -0.442 -1.474 -1.603 -1.650 175 serine/threonine kinase 6 pseudogene 540912 -1.043 -1.241 -1.045 0.184 176 growth arrest-specific 1 142146 -0.963 -1.757 -0.821 -0.315 177 Rho family GTPase 3 129829 1.101 -0.424 -0.641 -0.702 178 CDC20 cell division cycle 20 homolog (S. cerevisiae) 106187 -0.838 -0.769 -0.602 0.830 179 homocysteine-inducible, endoplasmic reticulum stress-inducible, 183516 1.591 0.216 0.259 -0.006 ubiquitin-like domain member 1 180 lymphotoxin beta (TNF superfamily, member 3) 180671 -0.783 -1.163 -1.890 -1.534 181 N/A 164460 0.167 1.337 1.564 0.056 168 LEC stimulated with VEGF-A Log2 Ratio MB Ranking Name Probe ID 1 hr 4 hr. 8 hr. 24 hr. 182 serum deprivation response (phosphatidylserine binding protein) 156433 0.391 -0.872 -1.088 -1.016 183 bone morphogenetic protein 2 193689 2.015 1.237 1.382 0.480 184 cell division cycle associated 8 130297 -0.859 -0.746 -0.550 0.756 185 cingulin-like 1 106427 -0.077 -1.020 -1.827 -1.774 186 shugoshin-like 1 (S. pombe) 227151 -0.686 -1.152 -0.710 1.185 187 cell division cycle associated 5 135130 -0.348 -0.454 0.215 1.462 188 peptidylglycine alpha-amidating monooxygenase COOH-terminal 149253 -0.985 -1.389 -0.689 0.157 interactor 189 B-cell CLL/lymphoma 6, member B (zinc finger protein) 178868 1.178 -0.071 0.047 -0.324 190 N/A 106827 -0.202 1.736 0.586 0.540 191 F-box and leucine-rich repeat protein 20 147632 0.447 -0.928 -1.064 -0.145 192 inositol 1,3,4,5,6-pentakisphosphate 2-kinase 168375 1.596 0.550 0.897 0.092 193 chromosome 10 open reading frame 114 541266 -0.028 -1.229 -1.752 -0.667 194 Ras-related associated with diabetes 149158 2.635 0.929 0.074 -0.025 195 transforming growth factor, beta 3 142790 1.346 1.832 0.339 0.043 196 chromosome 8 open reading frame 4 108623 0.139 -2.079 -2.106 -0.871 197 CCAAT/enhancer binding protein (C/EBP), delta 151000 1.621 0.476 -0.319 0.198 198 sorting nexin 22 145714 1.255 1.908 1.373 1.203 199 tumor necrosis factor, alpha-induced protein 8-like 3 151078 1.727 0.555 -0.920 -1.080 200 interferon induced with helicase C domain 1 133193 -0.034 -1.480 -0.953 -0.546 201 centromere protein E, 312kDa 165425 0.168 0.057 0.179 1.811 202 N/A 123427 1.267 -0.031 -0.279 0.190 203 solute carrier family 25 (mitochondrial carrier; phosphate carrier), 183778 1.793 0.711 0.677 0.406 member 25 204 transforming growth factor, alpha 180395 0.022 -0.885 -1.195 -1.646 205 N/A 200017 0.710 1.617 1.957 0.841 206 diaphanous homolog 3 (Drosophila) 102085 -0.398 -0.434 -0.284 1.252 207 Fanconi anemia, complementation group D2 151336 -1.439 -1.088 -1.339 0.423 208 potassium intermediate/small conductance calcium-activated channel, 102973 0.653 2.303 1.067 1.972 subfamily N, member 2 209 N/A 119241 0.207 -0.234 -0.205 1.343 210 ZW10 interactor 167013 -0.229 -0.277 0.156 1.494 211 p300/CBP-associated factor 183086 -0.084 -0.999 -1.729 -0.234 212 maternal embryonic leucine zipper kinase 156792 -0.332 -0.448 0.158 1.266 213 FXYD domain containing ion transport regulator 3 201258 0.934 -0.065 1.172 1.164 214 H2.0-like homeobox 1 (Drosophila) 197242 2.067 0.912 0.554 1.123 215 fibronectin leucine rich transmembrane protein 2 171313 0.241 -0.902 -0.938 -1.258 216 interferon-induced protein with tetratricopeptide repeats 2 130677 0.567 -1.331 -1.305 -0.478 217 thyroid hormone receptor interactor 13 194658 -0.799 -0.294 0.297 0.950 218 FSH primary response (LRPR1 homolog, rat) 1 187170 -0.385 -0.928 0.516 1.020 219 cAMP responsive element modulator 141393 1.269 0.673 -0.609 -0.221 220 cell division cycle associated 2 166387 -1.104 -0.481 -0.714 0.768 221 calcium/calmodulin-dependent protein kinase kinase 1, alpha 150074 -0.186 1.307 0.544 -0.041 222 kinesin family member 14 121673 0.183 -0.287 -0.060 1.792 223 neuronal PAS domain protein 2 185464 0.376 1.384 1.648 0.615 224 chemokine (C-C motif) ligand 20 133100 1.672 1.012 -0.689 -0.299 225 chromosome 1 open reading frame 21 185996 -0.010 1.436 0.772 0.651 226 histone 1, H4d 112668 -0.466 -0.594 -0.345 1.040 227 N/A 144657 0.117 -0.306 -1.379 -1.229 169 LEC stimulated with VEGF-A Log2 Ratio MB Ranking Name Probe ID 1 hr 4 hr. 8 hr. 24 hr. 228 cache domain containing 1 214499 0.397 -0.911 -1.186 -0.686 229 breast cancer 2, early onset 119580 0.272 -0.510 0.585 1.431 230 chromosome 10 open reading frame 10 178105 -0.657 0.739 0.395 0.625 231 pituitary tumor-transforming 2 130575 -0.557 -0.649 -0.895 0.680 232 high-mobility group box 2 177404 -0.732 -1.221 -0.830 0.448 233 insulin-like growth factor 1 (somatomedin C) 123381 0.489 -1.035 -1.462 -1.389 234 family with sequence similarity 54, member A 143413 -0.681 -0.255 0.285 0.950 235 N/A 144215 -0.077 1.193 0.589 -0.406 236 programmed cell death 4 (neoplastic transformation inhibitor) 144148 0.322 -0.611 -1.699 -0.838 237 KIAA0101 123361 -0.240 -0.524 -0.621 1.154 238 kelch-like 24 (Drosophila) 188332 0.321 -0.450 -1.241 0.240 239 FOS-like antigen 2 117028 0.859 -0.071 0.035 -0.831 240 cytochrome P450, family 4, subfamily X, polypeptide 1 123084 -0.247 -0.658 -1.494 -1.370 241 leucine-rich repeats and immunoglobulin-like domains 3 137001 1.365 -0.845 -0.465 -0.294 242 M-phase phosphoprotein 1 205220 0.042 -0.159 0.440 1.495 243 galanin 167511 -0.441 1.466 1.819 0.900 244 diaphanous homolog 3 (Drosophila) 108052 -0.330 -0.592 -0.402 1.411 245 interferon stimulated exonuclease gene 20kDa 118954 0.547 1.520 1.411 0.385 246 KIAA0513 104925 1.124 0.721 -0.654 -0.511 247 N/A 187106 0.293 0.193 -0.380 -1.496 248 kelch-like 4 (Drosophila) 114429 0.400 0.670 -0.590 -0.707 249 Kruppel-like factor 2 (lung) 174556 1.730 0.392 -0.742 -0.198 250 KIAA1913 229140 0.793 2.200 0.495 0.402 251 polo-like kinase 4 (Drosophila) 191973 0.125 -0.471 0.113 1.298 252 AXIN1 up-regulated 1 151956 3.185 1.228 0.922 0.440 253 nuclear receptor subfamily 4, group A, member 3 102081 3.668 0.750 0.310 0.424 254 syndecan 2 (heparan sulfate proteoglycan 1, cell surface-associated, 209676 -0.117 -0.449 -1.174 -1.422 fibroglycan) 255 N/A 182019 -0.342 -0.695 -1.629 -1.191 256 solute carrier family 45, member 4 222162 1.177 0.694 -0.731 0.094 257 fatty acid binding protein 3, muscle and heart (mammary-derived 146250 0.486 1.003 1.446 2.424 growth inhibitor) 258 v-myb myeloblastosis viral oncogene homolog (avian)-like 2 154410 -0.550 -0.827 -0.161 1.115 259 polymerase (RNA) I polypeptide B, 128kDa 219437 -0.624 0.729 0.525 -0.381 260 hyaluronoglucosaminidase 4 189554 0.633 0.386 0.019 1.525 261 myeloid/lymphoid or mixed-lineage leukemia (trithorax homolog, 129350 0.066 -1.029 -1.394 -0.869 Drosophila); translocated to, 3 262 glycine dehydrogenase (decarboxylating; glycine decarboxylase, 114924 -1.207 -1.407 -0.895 -0.292 glycine cleavage system protein P) 263 matrix metallopeptidase 19 708133 -0.141 -0.806 0.001 -1.341 264 leucine zipper protein 5 212326 -0.573 -0.698 -0.613 0.903 265 chromosome 12 open reading frame 24 214857 -0.302 0.861 1.317 -0.096 266 Bloom syndrome 199557 -0.047 -0.126 0.749 1.403 267 KIAA1370 186024 0.608 -0.625 -0.090 0.764 268 N/A 182347 -0.423 0.158 -1.147 -1.147 269 IQ motif containing GTPase activating protein 3 186522 -0.738 -0.923 -0.386 0.756 270 kinesin family member 15 188149 -0.367 -0.384 -0.180 1.259 271 pseudouridylate synthase 1 202211 -0.154 0.951 1.112 -0.023 272 chromosome 20 open reading frame 42 172479 -0.570 0.872 1.091 2.451 273 zinc finger protein 664 187860 -1.346 -1.122 -1.702 -1.286 170 LEC stimulated with VEGF-A Log2 Ratio MB Ranking Name Probe ID 1 hr 4 hr. 8 hr. 24 hr. 274 CHK1 checkpoint homolog (S. pombe) 211140 -0.763 -0.908 0.354 0.492 275 N/A 179063 -0.454 -0.790 -0.599 0.933 276 dedicator of cytokinesis 10 107398 -0.444 1.129 0.624 -0.108 277 ADAM metallopeptidase with thrombospondin type 1 motif, 18 220286 1.239 0.343 0.032 -0.425 278 decay accelerating factor for complement (CD55, Cromer blood group 167208 0.620 1.662 1.701 1.651 system) 279 l(3)mbt-like (Drosophila) 115601 0.806 -0.827 -0.439 -0.289 280 TAF4b RNA polymerase II, TATA box binding protein (TBP)- 127808 0.084 1.530 0.599 0.323 associated factor, 105kDa 281 G protein-coupled receptor 125 236580 -0.923 0.276 -0.750 -0.988 282 ankyrin repeat domain 20B 161943 0.751 1.645 1.807 0.955 283 potassium channel tetramerisation domain containing 12 105004 0.096 -0.968 -2.305 -2.036 284 tissue factor pathway inhibitor 2 144600 0.207 0.965 1.285 1.261 285 phospholipase C, beta 4 100736 0.735 -0.332 -0.858 -0.479 286 acyl-CoA thioesterase 11 164864 0.031 1.338 1.805 1.203 287 WD repeats and SOF1 domain containing 196524 1.784 1.456 1.510 0.861 288 RAD51 homolog (RecA homolog, E. coli) (S. cerevisiae) 158651 -0.464 -0.242 0.527 1.183 289 dual specificity phosphatase 26 (putative) 185108 -0.468 -1.121 -0.786 0.635 290 chromosome 9 open reading frame 76 177183 -0.393 -0.767 0.250 0.730 291 N/A 236648 1.844 0.942 0.857 0.526 292 decay accelerating factor for complement (CD55, Cromer blood group 109858 0.598 1.262 1.023 1.600 system) 293 carbohydrate (keratan sulfate Gal-6) sulfotransferase 1 151863 -0.218 -0.775 0.324 1.025 294 chromosome 7 open reading frame 31 172392 -0.280 -1.519 -1.114 -0.515 295 plasminogen activator, urokinase receptor 208060 0.848 2.362 2.244 1.423 296 CDC28 protein kinase regulatory subunit 1B 112634 -0.509 -0.461 -0.451 0.852 297 ubiquitin-conjugating enzyme E2T (putative) 208303 -0.276 -0.282 0.260 1.265 298 superoxide dismutase 2, mitochondrial 149133 0.937 1.328 1.353 1.534 299 shugoshin-like 2 (S. pombe) 215796 0.126 -0.463 0.012 1.131 300 Ras association (RalGDS/AF-6) domain family 8 164634 0.465 0.344 -0.184 -1.082 301 ADP-ribosylation factor-like 5B 182434 1.422 0.386 0.469 0.567 302 BCL6 co-repressor 223980 1.843 0.749 0.575 0.383 303 tumor necrosis factor (ligand) superfamily, member 10 153282 0.060 -1.419 -1.306 -0.475 304 ASF1 anti-silencing function 1 homolog B (S. cerevisiae) 161404 -0.433 -0.748 -0.009 1.021 305 breast cancer 1, early onset 124112 -0.262 -0.711 0.286 1.378 306 N/A 190904 -0.216 -0.566 -0.936 1.136 307 diacylglycerol kinase, delta 130kDa 194946 0.743 1.213 0.623 1.635 308 dihydrofolate reductase 114480 -0.459 -0.480 -0.371 0.896 309 purinergic receptor P2Y, G-protein coupled, 5 189269 0.682 -0.565 -0.704 0.262 310 nucleolar protein 5A (56kDa with KKE/D repeat) 189611 -0.089 1.276 0.936 0.191 311 chromosome 6 open reading frame 173 228202 -0.782 -0.807 -0.590 0.581 312 plasminogen activator, tissue 152725 0.385 2.268 1.835 -0.094 313 chromosome 1 open reading frame 51 190300 0.768 -0.622 -1.350 -0.758 314 N/A 716249 -0.336 -1.387 -0.329 -0.692 315 cytoplasmic polyadenylation element binding protein 4 202878 1.574 0.451 0.476 0.505 316 v-maf musculoaponeurotic fibrosarcoma oncogene homolog F (avian) 179935 1.548 0.200 -0.085 -0.135 317 SERTA domain containing 1 183615 2.051 0.518 0.331 0.180 318 chromosome 9 open reading frame 68 126803 0.832 -0.554 -0.735 -0.094 319 RALBP1 associated Eps domain containing 2 125119 0.038 -0.602 -1.691 -1.072 171 LEC stimulated with VEGF-A Log2 Ratio MB Ranking Name Probe ID 1 hr 4 hr. 8 hr. 24 hr. 320 ELOVL family member 6, elongation of long chain fatty acids 216339 -0.088 0.476 1.246 0.987 (FEN1/Elo2, SUR4/Elo3-like, yeast) 321 testis-specific kinase 2 193067 -0.169 -1.151 -1.413 -0.174 322 N/A 212060 0.651 0.184 -0.906 0.249 323 E2F transcription factor 8 185998 -0.650 -0.427 -0.503 0.823 324 immediate early response 2 163612 2.048 0.846 0.654 0.455 325 N/A 109626 -0.139 -0.963 0.423 0.296 326 single-stranded DNA binding protein 2 182329 -0.064 0.374 -1.034 -0.875 327 DnaJ (Hsp40) homolog, subfamily B, member 4 103618 -0.108 -1.090 -1.386 -1.303 328 N/A 100458 -0.406 -0.189 1.244 0.610 329 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3 213278 1.618 1.337 1.085 0.573 330 N/A 212778 -0.419 -0.448 -0.496 0.889 331 ras homolog gene family, member J 168924 0.097 0.152 -1.174 -1.221 332 synaptogyrin 3 147437 0.228 1.046 0.371 -0.462 333 mannosidase, alpha, class 1C, member 1 211828 -0.033 -0.499 -0.821 -1.521 334 regulator of G-protein signalling 2, 24kDa 116793 2.021 -0.396 -0.622 -0.552 335 synapse defective 1, Rho GTPase, homolog 1 (C. elegans) 104549 -1.572 -0.754 -0.502 -0.676 336 Fanconi anemia, complementation group B 141313 -0.125 -0.622 0.423 0.812 337 tribbles homolog 1 (Drosophila) 150749 3.576 1.384 0.629 0.148 338 low density lipoprotein receptor-related protein 8, apolipoprotein e 221017 -0.052 1.371 1.880 1.195 receptor 339 ral guanine nucleotide dissociation stimulator-like 1 180843 1.352 1.298 1.210 0.607 340 chromosome 2 open reading frame 23 156624 -0.222 -1.115 -1.807 -1.107 341 nuclear receptor subfamily 1, group H, member 3 155957 -0.041 -1.067 -0.972 -1.142 342 bruno-like 5, RNA binding protein (Drosophila) 193658 -0.135 -0.924 -1.283 -1.663 343 N/A 111796 0.185 -0.097 -0.479 -1.290 344 KIAA1794 190099 -0.247 -0.386 0.066 1.325 345 ATPase family, AAA domain containing 3C 148869 -1.012 0.119 0.291 -0.526 346 chromosome 9 open reading frame 95 146066 0.240 -0.807 -1.113 -0.355 347 KIAA1914 129030 -0.072 -1.765 -1.637 -1.444 348 ADAM metallopeptidase with thrombospondin type 1 motif, 4 212524 1.173 0.632 0.504 1.441 349 interferon-related developmental regulator 2 213763 -0.597 0.319 0.873 0.133 350 trans-prenyltransferase 117679 -0.647 0.960 1.397 0.447 351 nucleolar protein family 6 (RNA-associated) 224105 -0.177 1.001 1.067 0.055 352 chromosome 14 open reading frame 145 154359 -0.393 -0.960 -0.416 0.606 353 proline-rich nuclear receptor coactivator 1 102937 0.289 -0.415 -1.110 0.065 354 solute carrier family 40 (iron-regulated transporter), member 1 165911 -0.094 -1.175 -1.151 -0.539 355 AT rich interactive domain 5B (MRF1-like) 163382 1.258 -0.064 -0.016 0.231 356 growth arrest-specific 2 like 3 137837 0.421 -0.792 -0.521 0.530 357 transmembrane protein 88 200951 1.148 -0.252 -0.498 -0.688 358 chromosome 20 open reading frame 19 213234 0.176 -0.725 -1.440 -0.389 359 phosphatidic acid phosphatase type 2C 120143 0.343 0.274 -0.464 -1.080 360 N/A 194097 0.237 -0.347 -1.197 -0.652 361 N/A 228136 0.400 -0.760 -0.843 -0.594 362 CD200 antigen 105368 1.281 0.185 -0.190 0.134 363 N/A 208543 -0.149 -0.845 -0.939 -1.469 364 N/A 197968 -0.004 -0.593 -1.479 -0.532 365 ring finger protein 150 213739 0.536 -0.398 -0.617 -0.707 366 N/A 123821 -0.326 0.914 0.385 0.728 172 LEC stimulated with VEGF-A Log2 Ratio MB Ranking Name Probe ID 1 hr 4 hr. 8 hr. 24 hr. 367 angiopoietin 2 193875 0.459 1.960 2.538 2.302 368 platelet derived growth factor C 189538 0.258 -0.440 -1.631 -1.078 369 RNA, U3 small nucleolar interacting protein 2 206116 -0.252 1.164 1.071 -0.074 370 translocase of outer mitochondrial membrane 40 homolog (yeast) 197942 -0.312 0.622 1.022 0.058 371 N/A 222424 1.014 -0.447 0.040 0.008 372 PDZ and LIM domain 4 170266 -0.120 1.400 0.736 0.204 373 Kruppel-like factor 5 (intestinal) 156864 1.531 1.405 0.768 0.801 374 v-jun sarcoma virus 17 oncogene homolog (avian) 123273 0.680 0.214 -0.654 -0.869 375 protein-L-isoaspartate (D-aspartate) O-methyltransferase domain 129701 0.064 -0.932 -1.551 -0.435 containing 1 376 ST8 alpha-N-acetyl-neuraminide alpha-2,8-sialyltransferase 4 203385 0.577 0.852 0.462 1.842 377 signal transducer and activator of transcription 1, 91kDa 136002 0.105 -0.299 -1.175 -1.054 378 family with sequence similarity 64, member A 159220 -0.678 -0.423 -0.607 0.857 379 thymidylate synthetase 154415 -0.535 -0.762 -0.382 0.978 380 syntaxin 11 138768 1.237 1.059 0.561 0.027 381 zinc finger protein 323 199498 -0.548 -1.528 -1.690 -0.947 382 lunatic fringe homolog (Drosophila) 185842 -0.548 -1.205 -1.561 -1.473 383 iroquois homeobox protein 2 542107 1.811 0.531 0.258 0.245 384 RAD54 homolog B (S. cerevisiae) 115487 -0.535 -0.973 0.242 0.435 385 N/A 297413 1.511 -0.139 0.398 0.461 386 N/A 195134 -0.013 -0.138 0.346 1.469 387 ubiquitin-like, containing PHD and RING finger domains, 1 165886 -0.426 0.059 1.564 0.943 388 ring finger protein 144 101012 0.028 -1.448 -0.960 -0.841 389 coiled-coil domain containing 33 235214 0.275 -0.758 -0.085 0.557 390 peter pan homolog (Drosophila) 171791 -0.064 1.068 0.934 -0.159 391 integrin, alpha 7 185859 1.094 0.402 0.460 1.326 392 N/A 151231 1.059 -0.226 -0.036 0.990 393 desmuslin 114901 0.512 1.293 1.169 1.087 394 cell division cycle associated 7 207416 -0.198 0.856 1.190 0.045 395 hydroxysteroid dehydrogenase like 1 192606 -1.483 -0.824 -1.109 -1.230 396 apolipoprotein L, 4 147557 -0.320 -1.118 -1.958 -0.841 397 copine family member IX 214846 0.244 -0.082 -0.478 -1.135 398 CDC6 cell division cycle 6 homolog (S. cerevisiae) 130465 -0.193 -0.030 1.303 1.230 399 pecanex homolog (Drosophila) 134496 0.100 -0.838 -1.119 -0.724 400 N/A 206048 -0.048 -1.307 -0.874 -0.638 173 Appendix Table 3 Top 400 differentially modulated genes by VEGF-C in LEC using Multivariate Bayesian ranking analysis LEC stimulated with VEGF-C Log2 Ratio MB Ranking Name Probe ID 1 hr 4 hr. 8 hr. 24 hr. 1 v-fos FBJ murine osteosarcoma viral oncogene homolog 205128 3.255 0.995 1.516 2.035 2 early growth response 1 147353 3.731 0.563 1.185 1.358 3 early growth response 3 124744 5.491 0.119 0.460 0.708 4 coagulation factor III (thromboplastin, tissue factor) 146916 2.830 1.368 0.243 1.063 5 prostaglandin-endoperoxide synthase 2 (prostaglandin G/H 208388 2.193 0.046 0.712 0.376 synthase and cyclooxygenase) 6 N/A 133024 -0.499 -1.841 -1.669 -1.644 7 early growth response 2 (Krox-20 homolog, Drosophila) 101929 4.604 -0.018 0.123 0.302 8 N/A 228964 -0.628 0.259 -1.974 -0.890 9 ADAM metallopeptidase with thrombospondin type 1 motif, 9 160876 1.965 0.669 -0.061 0.461 10 coagulation factor III (thromboplastin, tissue factor) 204787 3.240 1.563 0.711 1.095 11 p300/CBP-associated factor 183086 0.296 -0.574 -1.632 -0.487 12 N/A 178581 1.895 0.790 0.512 1.008 13 SNF1-like kinase 171526 2.100 0.462 0.547 0.356 14 zinc finger protein 36, C3H type, homolog (mouse) 179827 1.839 0.757 0.763 1.148 15 DEP domain containing 1B 206865 -0.648 -1.147 -1.255 0.259 16 protein regulator of cytokinesis 1 180626 -0.627 -0.721 -0.597 0.737 17 endothelial cell-specific molecule 1 174810 0.263 1.655 1.375 1.738 18 guanylate cyclase 1, soluble, alpha 3 170165 0.792 1.103 -0.385 -0.529 19 polo-like kinase 1 (Drosophila) 197341 -0.693 -0.867 -1.531 0.208 20 aurora kinase B 203163 -0.716 -0.931 -0.565 0.567 21 phosphoinositide-3-kinase, regulatory subunit 3 (p55, gamma) 216089 0.580 -0.944 -0.694 -0.186 22 KIAA1913 229140 0.691 1.569 -0.019 0.323 23 jumonji domain containing 3 124424 0.845 -0.459 -0.561 0.111 24 nuclear receptor subfamily 4, group A, member 1 216600 3.591 0.546 0.510 0.305 25 N/A 185597 -0.294 1.217 0.603 0.261 26 interleukin 1, beta 130322 1.674 0.759 0.389 0.854 27 lymphotoxin beta (TNF superfamily, member 3) 180671 -0.754 -0.910 -1.481 -1.483 28 latrophilin 1 127718 0.237 -0.684 0.909 0.931 29 midnolin 159555 1.809 0.596 -0.101 0.359 30 NIMA (never in mitosis gene a)-related kinase 2 115004 -0.777 -1.255 -1.057 0.071 31 cyclin-dependent kinase inhibitor 2C (p18, inhibits CDK4) 112997 -0.708 -1.589 -1.175 -0.230 32 kinesin family member 4A 115354 -0.687 -0.833 -0.986 0.409 33 chromosome 15 open reading frame 42 196613 -0.508 -0.734 -0.215 0.722 34 kinesin family member 20A 118830 -1.165 -1.511 -0.966 -0.272 35 zinc finger protein 614 147595 1.413 1.157 0.526 0.656 36 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3 198714 1.539 0.950 0.309 0.856 37 iroquois homeobox protein 2 542107 1.430 0.511 0.036 0.179 38 SERTA domain containing 1 183615 1.292 0.023 0.145 0.053 39 thymidine kinase 1, soluble 105119 -0.672 -0.747 -0.815 0.448 40 calmin (calponin-like, transmembrane) 162511 -0.787 0.477 -0.800 -0.745 41 N/A 209213 1.349 -0.719 -0.589 -0.587 174 LEC stimulated with VEGF-C Log2 Ratio MB Ranking Name Probe ID 1 hr 4 hr. 8 hr. 24 hr. 42 N/A 191516 -1.004 -1.002 -0.157 -1.056 43 N/A 230575 0.814 0.877 -0.224 -0.196 44 N/A 210759 -0.408 0.734 -0.079 -0.873 45 apolipoprotein B mRNA editing enzyme, catalytic polypeptide-like 138620 -0.892 -1.347 -0.956 -0.167 3B 46 N/A 179136 1.263 -0.252 0.752 0.550 47 Cbp/p300-interacting transactivator, with Glu/Asp-rich carboxy- 118816 -0.722 0.172 -1.567 -1.696 terminal domain, 1 48 zinc finger protein 664 187860 -1.135 -0.870 -1.731 -0.995 49 cyclin B2 169571 -0.645 -1.069 -1.038 0.125 50 nucleolar and spindle associated protein 1 128435 -0.404 -0.638 -0.706 0.754 51 fatty acid binding protein 3, muscle and heart (mammary-derived 146250 0.646 0.592 1.193 1.677 growth inhibitor) 52 PDZ binding kinase 169723 -0.794 -1.040 -1.197 0.031 53 topoisomerase (DNA) II alpha 170kDa 135302 -0.264 -0.461 -0.430 0.820 54 chemokine orphan receptor 1 139192 0.877 1.014 -0.560 0.166 55 nitric oxide synthase 1 (neuronal) adaptor protein 180965 -0.573 -0.621 -0.571 0.532 56 baculoviral IAP repeat-containing 5 (survivin) 104062 -0.759 -1.065 -0.816 0.132 57 antigen identified by monoclonal antibody Ki-67 137656 -0.369 -0.573 -0.658 0.653 58 centromere protein F, 350/400ka (mitosin) 183726 -0.038 -0.358 -0.309 0.984 59 BCL6 co-repressor 223980 1.396 0.492 0.708 0.826 60 pyruvate dehydrogenase kinase, isozyme 4 101060 -1.226 -1.517 -1.566 -1.107 61 N/A 228086 0.424 -0.587 -0.740 -0.757 62 N/A 620594 1.178 1.238 0.455 0.880 63 phosphoinositide-3-kinase, regulatory subunit 1 (p85 alpha) 180157 1.059 0.190 -1.010 -0.378 64 kinesin family member C1 141343 -0.540 -1.004 -0.689 0.322 65 solute carrier family 45, member 4 222162 1.057 0.404 0.035 -0.062 66 U2AF homology motif (UHM) kinase 1 218514 1.205 0.758 0.082 0.409 67 N/A 111700 -0.345 -0.573 -0.218 0.739 68 basic helix-loop-helix domain containing, class B, 2 197538 2.672 1.042 0.791 0.821 69 N/A 144262 0.722 -0.570 1.060 1.001 70 laminin, alpha 2 (merosin, congenital muscular dystrophy) 149994 0.210 -0.230 -0.681 -1.093 71 WD and tetratricopeptide repeats 1 205123 -1.184 -0.782 -0.393 -1.027 72 B-cell CLL/lymphoma 6, member B (zinc finger protein) 178868 0.939 0.263 -0.781 -0.111 73 chromosome 2 open reading frame 23 156624 -0.090 -0.711 -1.486 -1.186 74 N/A 135436 1.225 0.265 0.564 0.722 75 t-complex 11 (mouse) like 2 193950 0.490 0.157 -0.962 -0.024 76 baculoviral IAP repeat-containing 5 (survivin) 227666 -0.804 -0.889 -1.044 0.062 77 TSC22 domain family, member 2 113125 1.437 0.464 0.408 0.353 78 inhibitor of DNA binding 2, dominant negative helix-loop-helix 129260 0.697 -0.801 -0.509 -0.219 protein 79 solute carrier family 4, sodium bicarbonate cotransporter, member 7 163733 1.253 1.800 0.095 0.280 80 N/A 232041 1.789 1.290 -0.154 0.492 81 chromosome 18 open reading frame 56 223997 -0.569 0.012 -1.326 -0.239 82 calcium/calmodulin-dependent protein kinase (CaM kinase) II beta 227829 -0.139 1.219 0.531 0.446 83 elastin (supravalvular aortic stenosis, Williams-Beuren syndrome) 187671 -0.097 -0.732 0.390 -0.728 84 asp (abnormal spindle)-like, microcephaly associated (Drosophila) 181685 0.190 -0.197 -0.071 1.127 85 N/A 200967 0.816 -0.693 -0.474 -0.349 86 N/A 194279 0.481 1.623 0.460 -0.181 87 ras homolog gene family, member H 198968 -0.746 0.371 -0.300 -0.708 175 LEC stimulated with VEGF-C Log2 Ratio MB Ranking Name Probe ID 1 hr 4 hr. 8 hr. 24 hr. 88 N/A 128294 -0.305 -1.077 -0.195 -0.965 89 R-spondin family, member 4 208221 -0.473 -0.234 0.774 -0.428 90 chromosome 15 open reading frame 2 197925 -1.093 -0.004 -0.043 -0.429 91 ribonucleotide reductase M2 polypeptide 134286 -0.418 -0.275 -0.697 0.667 92 interferon-induced protein with tetratricopeptide repeats 2 130677 0.265 -0.943 -0.741 -0.443 93 inactivation escape 1 206913 0.517 -0.239 -0.221 -0.793 94 olfactory receptor, family 51, subfamily T, member 1 152842 -0.541 -0.553 -0.426 0.623 95 ubiquitin-conjugating enzyme E2C 143651 -0.563 -0.805 -0.480 0.385 96 N/A 231275 -0.401 -0.970 -1.359 -1.122 97 kinetochore associated 2 107406 -0.549 -0.789 -0.743 0.311 98 N/A 209907 -0.177 -0.845 0.098 -0.817 99 tubby like protein 1 137311 -0.427 0.530 0.682 -0.263 100 N/A 697680 1.369 1.127 0.891 1.049 101 activating transcription factor 3 185687 2.676 1.017 0.317 0.850 102 zinc finger protein 80 (pT17) 105989 -0.691 -0.536 0.471 -0.539 103 chemokine (C-X-C motif) receptor 4 191821 -0.617 -0.079 0.433 0.569 104 nei endonuclease VIII-like 3 (E. coli) 172546 -0.340 -0.608 -0.751 0.456 105 N/A 180079 -0.379 -0.337 -0.675 -1.443 106 regulator of G-protein signalling 2, 24kDa 116793 0.343 -0.536 -0.613 -0.760 107 fibroblast growth factor receptor 1 (fms-related tyrosine kinase 2, 217844 -0.239 -0.776 -1.856 -1.826 Pfeiffer syndrome) 108 N/A 175055 0.764 -0.038 0.648 -0.405 109 WD repeats and SOF1 domain containing 196524 1.886 1.506 1.455 1.019 110 gon-4 homolog (C.elegans) 191928 0.892 -0.402 0.619 -0.003 111 N/A 116001 0.531 -0.059 -0.426 -0.742 112 p21 (CDKN1A)-activated kinase 3 217739 1.234 0.815 0.660 0.090 113 N/A 234260 -0.322 -0.241 0.828 -0.536 114 N/A 121962 0.371 -0.043 -1.140 -0.233 115 glutaminyl-peptide cyclotransferase (glutaminyl cyclase) 152127 1.211 0.432 -0.131 0.339 116 hairy and enhancer of split 1, (Drosophila) 176983 1.418 -0.151 -0.608 -0.185 117 hairy/enhancer-of-split related with YRPW motif 1 121998 0.607 -0.856 -0.921 -0.567 118 aquaporin 7 124473 1.447 0.468 0.748 0.475 119 N/A 445006 1.257 0.290 0.173 0.055 120 pelota homolog (Drosophila) 222926 1.219 0.372 -0.072 0.546 121 calmegin 100646 0.547 0.755 -0.396 -0.196 122 v-maf musculoaponeurotic fibrosarcoma oncogene homolog B 226336 1.235 -0.744 -0.915 -0.040 (avian) 123 vesicle-associated membrane protein 1 (synaptobrevin 1) 192759 -0.833 -0.394 0.516 -0.545 124 stanniocalcin 1 119453 1.693 -0.077 -1.126 -0.404 125 N/A 198371 -1.263 -0.805 -1.031 -0.360 126 UDP-N-acetyl-alpha-D-galactosamine:polypeptide N- 142557 0.408 -0.166 0.440 1.074 acetylgalactosaminyltransferase 8 (GalNAc-T8) 127 N/A 151076 -0.580 -0.895 -0.686 -1.322 128 kinesin family member 2C 212531 -0.213 -0.632 -0.580 0.653 129 protein geranylgeranyltransferase type I, beta subunit 177456 0.674 -0.568 -0.112 0.031 130 dual specificity phosphatase 5 121612 2.564 0.889 0.576 0.682 131 N/A 154042 0.506 1.577 -0.207 -0.390 132 SHC SH2-domain binding protein 1 106848 -0.463 -0.509 -0.422 0.543 133 discs, large homolog 7 (Drosophila) 194498 -0.382 -0.617 -0.817 0.483 176 LEC stimulated with VEGF-C Log2 Ratio MB Ranking Name Probe ID 1 hr 4 hr. 8 hr. 24 hr. 134 N/A 159435 -1.060 -0.657 0.145 -0.309 135 cell division cycle associated 1 147806 -0.260 -0.687 -0.717 0.402 136 N/A 159898 -0.543 -1.043 -0.672 0.120 137 N/A 701184 0.779 0.260 0.950 -0.090 138 PBX/knotted 1 homeobox 1 137640 0.554 0.178 -0.302 0.841 139 kelch repeat and BTB (POZ) domain containing 6 201279 -0.988 0.174 -0.690 -0.196 140 chromosome 10 open reading frame 114 541266 -0.269 -1.160 -1.140 -0.608 141 N/A 236648 1.482 0.335 0.203 0.420 142 cyclin A2 110863 -0.471 -0.624 -0.823 0.328 143 T-box 3 (ulnar mammary syndrome) 115993 0.985 0.332 -0.120 -0.090 144 transmembrane protein 100 216519 -0.382 -1.114 -0.737 -1.143 145 decay accelerating factor for complement (CD55, Cromer blood 167208 0.258 1.133 0.727 0.836 group system) 146 myosin IB 115188 1.464 1.054 0.992 0.824 147 cAMP responsive element binding protein 5 211684 0.865 1.104 0.352 1.002 148 SUMO1/sentrin specific peptidase 1 224778 0.574 0.079 -0.680 0.281 149 interleukin 18 receptor 1 213445 -0.072 1.337 0.137 -0.600 150 N/A 215658 0.855 1.048 0.933 1.271 151 N/A 289734 0.837 -0.312 0.469 0.070 152 N/A 169184 0.563 1.020 -0.137 0.193 153 ADAM metallopeptidase with thrombospondin type 1 motif, 9 10298121 0.848 0.167 0.413 -0.520 154 cytochrome P450, family 1, subfamily A, polypeptide 1 135086 0.301 -0.419 -1.085 -0.978 155 C-type lectin domain family 2, member L 184748 -0.983 -0.011 -0.233 0.041 156 N/A 107502 -0.329 -0.416 -0.212 0.810 157 N/A 231119 0.052 1.012 -0.001 -0.321 158 centromere protein A, 17kDa 106198 -0.835 -0.915 -0.807 0.034 159 cell division cycle 2, G1 to S and G2 to M 123400 -0.541 -0.536 -0.542 0.434 160 N/A 233248 -0.756 0.240 -0.223 0.352 161 N/A 707294 1.035 0.485 -0.036 0.179 162 N/A 215902 -0.135 -0.534 0.084 0.786 163 chromosome 14 open reading frame 4 216855 0.813 -0.028 -0.422 0.044 164 transcription elongation factor A (SII), 3 107155 -0.019 0.189 -0.034 -0.912 165 heparan sulfate (glucosamine) 3-O-sulfotransferase 1 154628 0.464 -0.268 -0.550 -0.569 166 mal, T-cell differentiation protein-like 201519 0.395 0.006 -0.530 -0.720 167 spermatogenesis associated 12 229375 -0.736 -1.057 -1.372 -0.884 168 peroxisome proliferative activated receptor, gamma, coactivator 1, 147539 0.878 0.104 0.707 0.986 alpha 169 cyclin-dependent kinase inhibitor 3 (CDK2-associated dual 191305 -0.703 -0.741 -0.683 0.345 specificity phosphatase) 170 WAS protein family, member 2 146329 -0.512 -0.963 -1.613 -1.092 171 N/A 113704 -0.148 -0.818 -0.465 -0.953 172 cell division cycle associated 5 135130 -0.319 -0.684 -0.480 0.546 173 chromosome 21 open reading frame 108 235773 0.996 0.363 -0.064 0.409 174 N/A 170978 -0.561 0.258 -0.729 -0.820 175 hyaluronan-mediated motility receptor (RHAMM) 216917 -0.371 -0.611 -0.673 0.520 176 chromosome 20 open reading frame 128 149623 0.981 0.915 1.040 1.087 177 chromosome 10 open reading frame 10 178105 -0.749 0.177 0.298 0.264 178 growth differentiation factor 15 182404 1.203 0.365 0.312 0.598 179 anillin, actin binding protein (scraps homolog, Drosophila) 169499 -0.200 -0.441 -0.515 0.576 177 LEC stimulated with VEGF-C Log2 Ratio MB Ranking Name Probe ID 1 hr 4 hr. 8 hr. 24 hr. 180 melanoma antigen family C, 2 160885 0.094 1.162 0.184 0.640 181 N/A 104707 -0.171 -1.333 -0.559 -1.263 182 gap junction protein, alpha 4, 37kDa (connexin 37) 202429 0.304 -0.311 -0.802 -0.753 183 serine/threonine kinase 6 157917 -0.736 -0.985 -0.583 0.124 184 vascular endothelial growth factor C 170337 0.252 1.486 0.650 1.278 185 KIAA0101 123361 -0.383 -0.689 -0.797 0.347 186 N/A 106597 -0.591 0.127 -0.067 -1.029 187 zinc finger and BTB domain containing 26 177467 0.490 -0.596 0.178 -0.355 188 serine/threonine kinase 6 pseudogene 540912 -1.090 -1.175 -0.957 -0.506 189 N/A 148170 -0.598 -0.189 0.434 0.671 190 chromosome 1 open reading frame 61 220684 -0.878 -0.690 -0.462 -1.218 191 N/A 111200 -0.244 -1.081 -0.736 -0.794 192 inositol 1,3,4,5,6-pentakisphosphate 2-kinase 168375 1.171 0.137 0.436 0.246 193 inhibitor of DNA binding 2B, dominant negative helix-loop-helix 233364 0.525 -0.373 -0.515 -0.590 protein 194 Fanconi anemia, complementation group D2 151336 -0.626 -0.844 -0.934 0.268 195 N/A 191465 -0.107 0.297 -0.012 1.262 196 protein phosphatase 1K (PP2C domain containing) 191909 -0.701 -0.157 0.208 0.397 197 TPX2, microtubule-associated, homolog (Xenopus laevis) 189094 -0.904 -0.925 -0.674 0.008 198 N/A 351669 0.922 -0.144 0.508 0.635 199 chromosome 18 open reading frame 58 173685 -0.372 -0.660 -1.354 0.043 200 family with sequence similarity 64, member A 201158 -1.002 -1.037 -0.552 0.039 201 N/A 231627 0.078 0.337 -0.213 1.127 202 forkhead box D1 108203 0.321 0.091 -1.150 -0.432 203 ankyrin repeat domain 20B 161943 0.920 1.665 1.219 0.841 204 N/A 118927 0.789 0.965 1.170 1.314 205 N/A 140753 0.953 0.493 0.118 -0.566 206 zinc finger protein 467 184463 0.330 -0.788 -0.621 -0.407 207 N/A 209417 -0.104 0.409 1.345 0.525 208 centrosomal protein 55kDa 198728 -0.445 -0.542 -0.615 0.617 209 spindle pole body component 25 homolog (S. cerevisiae) 130624 -0.623 -0.779 -0.720 0.192 210 Kruppel-like factor 5 (intestinal) 156864 0.987 0.913 0.239 0.437 211 TAP binding protein (tapasin) 143717 -0.070 0.312 -1.063 -0.620 212 N/A 171510 -0.968 0.113 -0.706 -0.469 213 N/A 400486 0.171 1.638 0.086 -0.103 214 deleted in liver cancer 1 207792 1.074 0.844 0.396 0.561 215 sperm associated antigen 5 185888 -0.666 -0.936 -0.891 0.012 216 kinesin family member 23 160577 -0.372 -0.258 -0.297 0.641 217 kinesin family member 14 121673 -0.118 -0.498 -0.225 0.844 218 AT rich interactive domain 3B (BRIGHT- like) 712742 0.997 0.058 0.136 0.306 219 growth arrest-specific 7 189097 0.828 -0.700 0.295 -0.004 220 N/A 128495 -0.244 -0.654 -0.314 0.555 221 N/A 230585 1.021 1.073 0.508 1.172 222 fatty acid synthase 154286 -0.456 0.362 -0.326 0.477 223 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3 213278 1.241 0.797 0.200 0.602 224 N/A 159940 -0.473 0.159 0.711 -0.448 225 zinc finger protein 671 149250 -1.004 -0.163 -0.427 0.281 226 N/A 704926 1.012 0.114 1.430 0.778 178 LEC stimulated with VEGF-C Log2 Ratio MB Ranking Name Probe ID 1 hr 4 hr. 8 hr. 24 hr. 227 zinc finger protein 407 214534 -0.671 -0.935 -0.381 -1.128 228 N/A 192010 0.733 0.591 0.748 1.286 229 galanin 167511 -0.678 0.084 0.463 -0.005 230 adiponutrin 197774 0.576 1.000 0.857 1.070 231 KIAA1193 188138 -0.362 -0.761 0.140 -0.732 232 protein tyrosine phosphatase, receptor type, E 541414 1.594 0.811 0.602 0.063 233 B-cell CLL/lymphoma 6 (zinc finger protein 51) 151724 1.593 0.725 -0.094 0.560 234 fructose-1,6-bisphosphatase 1 125158 0.399 -0.742 -0.438 0.208 235 N/A 234709 1.654 0.656 0.543 0.332 236 tigger transposable element derived 1 237079 0.012 1.208 0.819 0.185 237 N/A 113701 0.124 1.091 0.707 0.442 238 N/A 217029 1.672 0.528 0.187 0.776 239 family with sequence similarity 91, member A2 103214 0.970 1.107 0.483 0.759 240 N/A 198050 -0.812 -0.733 0.160 -0.514 241 ets variant gene 1 216791 -0.055 0.991 0.721 0.178 242 N/A 205276 0.200 -0.891 -0.442 -0.576 243 ADP-ribosylation factor-like 5B 182434 0.899 -0.132 0.306 0.595 244 ERO1-like beta (S. cerevisiae) 207998 0.984 0.057 0.032 0.322 245 N/A 232731 -0.779 0.058 -0.433 -0.837 246 high-mobility group box 2 216889 -0.680 -0.935 -0.834 -0.045 247 paraneoplastic antigen MA3 176304 -0.093 -0.147 1.078 0.491 248 dual specificity phosphatase 10 187011 0.890 -0.105 0.011 -0.247 249 UTP14, U3 small nucleolar ribonucleoprotein, homolog A (yeast) 113975 -0.401 -0.360 -0.731 -1.226 250 sodium channel, voltage-gated, type III, alpha 161435 0.610 -0.392 0.144 0.634 251 G protein-coupled receptor 133 102836 -0.649 -0.857 0.197 -0.073 252 chemokine (C-X-C motif) receptor 3 103589 -0.106 -0.978 0.211 -0.459 253 N/A 245704 0.676 -0.292 0.845 0.133 254 low density lipoprotein receptor class A domain containing 1 548189 0.200 0.504 1.093 1.043 255 N/A 232983 -0.514 -1.135 -0.044 -0.824 256 peptidoglycan recognition protein 2 117570 -0.920 -0.017 -1.544 -1.575 257 scavenger receptor class F, member 2 148950 -0.473 -0.679 -1.799 -1.494 258 insulin induced gene 1 215650 1.015 0.087 0.221 0.389 259 N/A 182347 -0.447 -0.538 -1.138 -1.092 260 myeloid/lymphoid or mixed-lineage leukemia 3|B melanoma 199679 1.033 0.395 0.255 0.244 antigen family, member 5|B melanoma antigen family, member 3 261 low density lipoprotein receptor (familial hypercholesterolemia) 103382 1.135 0.609 0.422 0.664 262 N/A 181035 -0.254 -0.445 -1.424 -0.268 263 N/A 182262 -0.797 -0.068 0.028 -0.734 264 lymphocyte antigen 6 complex, locus H 173179 -0.228 0.715 0.919 0.034 265 M-phase phosphoprotein 6 226998 0.123 0.379 -0.250 1.019 266 N/A 481671 -0.590 0.127 0.459 -0.606 267 fibroblast growth factor 4 (heparin secretory transforming protein 1, 116891 -0.596 0.081 -1.011 -0.822 Kaposi sarcoma oncogene) 268 N/A 148766 -0.687 -0.185 0.448 -0.375 269 trophinin associated protein (tastin) 137875 -0.821 -1.123 -0.713 -0.115 270 N/A 114047 0.921 0.102 0.048 -0.237 271 ribosomal protein S6 kinase, 90kDa, polypeptide 5 152363 0.538 -0.219 -0.717 0.258 272 N/A 194013 -0.517 -0.079 0.042 -0.887 179 LEC stimulated with VEGF-C Log2 Ratio MB Ranking Name Probe ID 1 hr 4 hr. 8 hr. 24 hr. 273 phosphoenolpyruvate carboxykinase 1 (soluble) 112194 -0.287 -0.542 0.218 0.689 274 Wilms tumor 1 associated protein 227679 0.390 0.494 -0.730 -0.178 275 N/A 143181 0.230 0.807 0.214 1.002 276 N/A 222628 -0.157 -0.578 0.558 -0.347 277 CDC45 cell division cycle 45-like (S. cerevisiae) 208272 -0.491 -0.160 -0.099 0.795 278 hepatic leukemia factor 118529 -0.492 -0.232 0.243 -0.792 279 N/A 110350 0.561 -0.108 -0.407 -0.382 280 CTAGE family, member 4 235777 0.594 0.956 1.190 0.913 281 N/A 229613 -0.667 -1.091 -0.169 -0.998 282 transmembrane channel-like 8 173591 -1.210 -1.729 -1.428 -1.317 283 KIAA0513 104925 0.501 0.010 -0.617 -0.306 284 gamma-aminobutyric acid (GABA) B receptor, 1 195643 1.111 0.973 0.690 0.523 285 WAP four-disulfide core domain 12 131425 -0.931 -1.094 -0.068 -0.647 286 zinc finger protein 516 108345 -0.355 -0.737 -1.491 -0.091 287 Rac GTPase activating protein 1 135746 -0.340 -0.594 -0.691 0.333 288 hydroxysteroid dehydrogenase like 1 192606 -1.094 -1.218 -1.043 -0.398 289 N/A 226409 0.942 0.439 -0.252 -0.085 290 N/A 200017 0.834 1.627 1.071 0.851 291 potassium voltage-gated channel, KQT-like subfamily, member 1 183314 -0.787 -0.670 -0.513 -1.077 292 N/A 183303 1.862 0.556 -0.105 0.386 293 major histocompatibility complex, class II, DO beta 172619 0.465 1.162 0.564 0.211 294 nucleoside phosphorylase 147282 0.815 1.043 0.805 1.051 295 MLF1 interacting protein 147135 -0.541 -0.602 -0.226 0.464 296 N/A 335545 0.123 -0.761 0.193 -0.409 297 N/A 207583 -0.027 -0.551 0.530 -0.671 298 N/A 142167 0.212 -0.683 0.620 -0.174 299 N/A 712826 0.612 0.212 -0.304 -0.562 300 N/A 228136 0.555 0.264 -1.032 0.054 301 l(3)mbt-like 4 (Drosophila) 172568 0.126 0.175 -0.932 -0.327 302 Kruppel-like factor 10 104402 1.722 0.236 0.008 0.168 303 centromere protein A, 17kDa 128411 -0.682 -0.641 -1.098 0.263 304 phospholipase C, beta 4 100736 0.983 0.401 0.041 -0.065 305 KIAA0892 170092 -0.527 -0.478 -1.169 0.123 306 purinergic receptor P2Y, G-protein coupled, 8 133766 0.289 0.111 0.251 -0.661 307 N/A 151825 -1.052 -0.343 0.132 0.076 308 RAD9 homolog A (S. pombe) 208053 0.511 -0.643 -0.162 -0.285 309 chromosome 18 open reading frame 24 205780 -0.412 -0.441 -0.407 0.495 310 RAB5B, member RAS oncogene family 182445 -0.727 -0.591 -1.411 -0.375 311 chromosome 3 open reading frame 32 184605 0.559 0.806 0.628 1.373 312 kinesin family member 11 199107 -0.189 -0.465 -0.451 0.641 313 N/A 232185 -0.486 -0.769 0.057 -0.791 314 N/A 129073 -0.542 -1.125 -0.091 -0.203 315 N/A 147103 0.422 0.097 -0.225 -0.947 316 zinc finger protein 710 141826 -0.407 -0.100 -1.095 -0.811 317 inositol polyphosphate-5-phosphatase, 75kDa 213370 1.263 0.890 0.290 -0.126 318 coronin 6 150942 -0.013 0.024 0.775 0.831 319 adrenergic, beta-2-, receptor, surface 127856 0.634 -0.019 -0.381 -0.273 180 LEC stimulated with VEGF-C Log2 Ratio MB Ranking Name Probe ID 1 hr 4 hr. 8 hr. 24 hr. 320 chromosome 14 open reading frame 43 212518 -0.034 0.362 -0.299 -0.814 321 protein kinase, AMP-activated, gamma 2 non-catalytic subunit 542533 0.186 -0.558 0.598 -0.397 322 leucine zipper, down-regulated in cancer 1-like 114319 -0.468 0.291 -0.704 -0.729 323 palladin, cytoskeletal associated protein 130541 0.339 0.971 -0.151 -0.126 324 N/A 179552 -0.876 -0.091 -0.832 -0.190 325 basic helix-loop-helix domain containing, class B, 3 213577 0.250 -0.585 -0.173 -0.626 326 pipecolic acid oxidase 158722 0.845 1.189 1.205 1.257 327 N/A 648566 -1.036 0.130 0.022 -0.477 328 angiotensin II receptor, type 1 105197 -0.246 -0.873 0.297 -0.382 329 chromosome 20 open reading frame 58 211587 -0.717 -0.176 0.254 -0.569 330 Rho GTPase activating protein 11A 117485 -0.103 -0.284 -0.089 0.789 331 N/A 124168 -0.053 -0.376 -0.447 -0.929 332 proline/serine-rich coiled-coil 1 181161 -1.174 -1.160 -0.728 -0.242 333 protogenin homolog (Gallus gallus) 384810 -0.304 0.534 0.130 0.538 334 N/A 190308 0.063 -0.368 -0.535 0.592 335 discoidin domain receptor family, member 1 173211 0.857 0.326 0.906 0.934 336 RALBP1 associated Eps domain containing 2 125119 0.168 -0.339 -1.021 -0.212 337 N/A 185713 0.488 0.372 0.547 1.192 338 N/A 231579 -0.256 -0.173 0.496 0.632 339 serpin peptidase inhibitor, clade B (ovalbumin), member 3 163547 -0.535 -0.212 -0.491 0.549 340 trinucleotide repeat containing 9 125060 -0.220 0.771 0.256 0.832 341 N/A 138397 -0.432 -0.313 -0.207 0.573 342 N/A 206508 0.898 -0.274 1.359 1.548 343 forkhead box I1 157201 0.431 0.238 0.894 -0.184 344 N/A 261736 1.111 0.655 0.675 0.680 345 chromosome 17 open reading frame 38 180545 -1.123 -0.266 -0.823 -0.270 346 N/A 220154 0.552 1.187 0.840 0.172 347 pellino homolog 1 (Drosophila) 113197 1.132 0.235 0.054 0.197 348 myosin VIIB 187521 -0.695 -0.284 0.368 -0.087 349 N/A 120178 -0.578 -0.074 0.285 0.467 350 chromosome 1 open reading frame 127 227197 0.766 0.222 0.638 -0.126 351 solute carrier family 25 (mitochondrial carrier; phosphate carrier), 183778 1.148 0.533 0.239 0.213 member 25 352 tumor necrosis factor (ligand) superfamily, member 18 179164 -0.796 0.242 -0.779 -0.388 353 ST8 alpha-N-acetyl-neuraminide alpha-2,8-sialyltransferase 4 203385 0.249 0.460 0.350 1.051 354 kelch repeat and BTB (POZ) domain containing 2 211514 1.003 0.216 0.179 0.361 355 N/A 671775 0.826 1.030 0.312 0.457 356 N/A 567077 -0.941 -0.890 -0.363 -1.280 357 ZW10 interactor 167013 -0.377 -0.410 -0.187 0.621 358 aquaporin 2 (collecting duct) 129761 0.950 -0.299 0.559 0.237 359 N/A 107957 -0.355 -1.097 -0.917 -0.436 360 paxillin 155086 1.143 -0.211 1.404 0.953 361 N/A 235001 0.737 0.060 0.004 0.745 362 SLIT-ROBO Rho GTPase activating protein 2 165965 -0.369 -0.196 -1.122 -0.024 363 zinc finger protein 354C 128406 -0.846 0.184 -0.314 -0.400 364 N/A 181889 -1.041 -0.638 -0.761 -0.130 365 N/A 171415 0.376 1.027 0.266 0.958 366 polymerase (DNA directed), epsilon 2 (p59 subunit) 217170 -0.775 -0.742 -0.378 0.065 181 LEC stimulated with VEGF-C Log2 Ratio MB Ranking Name Probe ID 1 hr 4 hr. 8 hr. 24 hr. 367 peptidylprolyl isomerase (cyclophilin)-like 6 115831 0.335 -0.177 -0.261 -0.716 368 N/A 713730 0.209 0.325 1.251 0.102 369 calmodulin binding transcription activator 1 103913 0.638 1.175 0.624 0.665 370 CDC14 cell division cycle 14 homolog A (S. cerevisiae) 219032 0.088 1.116 0.606 0.292 371 potassium intermediate/small conductance calcium-activated 102973 0.581 1.308 0.285 0.889 channel, subfamily N, member 2 372 N/A 232921 0.228 -0.125 1.124 0.185 373 ADAM metallopeptidase with thrombospondin type 1 motif, 2 224359 -0.837 0.250 -0.077 -0.565 374 dual specificity phosphatase 4 148153 0.855 0.896 0.272 0.226 375 bicaudal D homolog 2 (Drosophila) 207433 0.290 0.327 -0.767 -0.087 376 potassium channel tetramerisation domain containing 2 232722 -0.476 0.026 -1.333 -0.581 377 integrin beta 1 binding protein (melusin) 2 116020 0.512 1.173 0.809 1.080 378 WNK lysine deficient protein kinase 4 170756 -0.158 0.423 0.916 0.074 379 sprouty homolog 4 (Drosophila) 148545 0.684 0.685 0.024 0.725 380 N/A 232482 -0.932 -0.072 -0.257 -0.838 381 double homeobox, 4 177482 0.420 -0.624 0.232 0.300 382 N/A 233420 0.191 0.733 0.744 -0.162 383 polycomb group ring finger 4 167637 -0.042 -0.231 -1.134 0.248 384 chromosome 9 open reading frame 41 205667 -0.960 -0.118 -0.518 -0.730 385 solute carrier family 2 (facilitated glucose transporter), member 3 180998 0.695 -0.131 -0.287 0.063 386 ependymin related protein 1 (zebrafish) 210940 0.623 0.838 -0.021 0.656 387 N/A 248285 0.418 0.082 0.645 -0.350 388 N/A 10409147 1.054 0.386 0.107 0.359 389 N/A 102342 0.747 -0.193 -0.070 -0.014 390 MKL/myocardin-like 2 112995 -0.969 -0.315 -0.302 -0.197 391 vestigial like 2 (Drosophila) 171918 -0.820 0.240 0.001 -0.493 392 LIM domain only 6 136580 -0.456 -0.322 -0.020 -0.945 393 CDC14 cell division cycle 14 homolog A (S. cerevisiae) 176146 0.240 1.107 0.171 0.319 394 serine/threonine kinase 6 pseudogene 683679 -0.567 -0.680 -0.520 0.231 395 N/A 192975 -0.420 -0.710 -0.910 0.124 396 FBJ murine osteosarcoma viral oncogene homolog B 105390 1.288 0.002 0.424 0.040 397 TTK protein kinase 107112 -0.346 -0.587 -0.534 0.422 398 N/A 228432 -1.001 -0.278 -0.023 -0.083 399 cystathionase (cystathionine gamma-lyase) 105377 -0.953 -0.334 -0.044 -0.153 400 solute carrier family 6, member 16 191984 0.961 0.518 0.521 0.109 182 Appendix Table 4 Pathway classification analysis of VEGF-C induced genes in LEC Molecular function 1h 4h 8h 24h Transcription factor +++ - - - cytokine receptor ++ ++ + +++ Zinc finger transcription factor ++ - - - Nucleic acid binding ++ - - - Kinase modulator ++ - - - Phosphatase ++ - - - Nuclear hormone receptor ++ - - - Carbohydrate phosphatase ++ - - - Cytokine receptor + +++ - ++ Protein phosphatase + + - - Carbohydrate transporter + - - - Kinase inhibitor + - - - Defense/immunity protein - +++ - - Phosphorylase - ++ - +++ Complement component - ++ - - Growth factor - + - + Lipase - + - + Immunoglobulin receptor family member - + - - Interleukin receptor - + - - Guanylate cyclase - + - - Storage protein - - + - Ligand-gated ion channel - - - + Glycosyltransferase - - - + Biological process 1h 4h 8h 24h MAPKKK cascade +++ ++ - - Nucleoside, nucleotide and nucleic acid metabolism +++ + - - Cell cycle control +++ - - - mRNA transcription +++ - - - mRNA transcription regulation +++ - - - Intracellular signaling cascade ++ ++ - - Cell cycle ++ + - - Developmental processes ++ + - - Monosaccharide metabolism ++ - - - Cell communication + + - - Ligand-mediated signaling + + - - Signal transduction + + - - Protein phosphorylation + + - - Oncogene + - + - Cell proliferation and differentiation + - - - JNK cascade + - - - Regulation of phosphate metabolism + - - - Complement-mediated immunity - ++ - - Angiogenesis - + - + Purine metabolism - + - + Ion transport - + - - Protein modification - + - - Fatty acid metabolism - + - - Cholesterol metabolism - - + - Blood clotting - - - + Synaptic transmission - - - + +++ p-value < 0.0005; ++ p-value < 0.005; + p-value < 0.05; - not significant 183 8 CURRICULUM VITAE Jay Woo Shin PERSONAL INFORMATION Date of Birth April 4, 1981 Place of Birth Seoul, South Korea Nationality United States of America Sex Male Marital status Single Address Regensdorferstrasse 9 8049 Zurich, Switzerland Phone: ++41 44 633 7368 Mobile: ++41 76 429 1037 [email protected] EDUCATION Ph.D. in Natural Sciences Apr 2008 Swiss Federal Institute of Technology (ETH) Zurich, Switzerland • Thesis: Identification of the vascular lineage-specific transcriptome and development of a novel low-density microvascular differentiation array. • Advisor: Prof. Dr. Michael Detmar Bachelor of Science in Computer Informatics May 2003 Boston College, Boston, MA, USA • Pre-medical program RESEARCH EXPERIENCES Research Student, Department of Biological Information Aug 2007 - Oct 2007 Tokyo Institute of Technology, Yokohama, Japan • Research topic: Cell-specific delivery of tumor suppressor gene using polyethylimine-complexed protein A. • Advisor: Prof. Eiry Kobatake Research Assistant, Cutaneous Biology Research Center Sep 2002 - Oct 2004 Massachusetts General Hospital/Harvard Medical School, Boston, MA, USA • Research topics: Investigation of FGF signaling in lymphangiogenesis and lymphatic re-programming of blood vascular endothelium by Prox1 regulation. • Advisor: Prof. Dr. Michael Detmar 184 AWARDS & SCHOLARSHIPS • ETH Scholarship to Tokyo Institute of Technology, Japan (2007) • Young Investigator Scholarship, Gordon Conference (2006) • European Molecular Biology Organization (EMBO) Scholarship (2006) • Young Investigator Award, Gene Signature Symposium. Zurich, Switzerland (2005) ACTIVITIES AND SKILLS • President of Pharmaceutical Scientists’ Association, PSA (2006, 2007) • Information Technology (IT) Officer of Pharmacogenomics (2004 - present) • Programming in C, Java, R, Perl, Python and Bioconductor implementation. • Fluent in English and Korean. Basic in Japanese and German. ORAL PRESENTATIONS 1. Shin J. Vascular endothelial growth factor-A mediates endocan to promote lymphangiogenesis. 3rd Cancer Network of Zurich Symposium. Fiesch, Switzerland (January 2008). 2. Shin J (Keynote speaker). Investigation of the blood and lymph vessels growth in cancer and inflammation. 1st Young Scientists Symposium. Munich, Germany (July 2007). 3. Shin J. Identification of the vascular lineage-specific transcriptome and development of a novel low-density microvascular differentiation array. 3rd International qPCR Symposium. Freising-Weihenstephan, Germany (March 2007). 4. Shin J. Comprehensive identification of vascular lineage-specific gene signatures and establishment of a low-density microvascular differentiation assay. RNA Symposium, Basel, Switzerland (November 2006). 5. Shin J. Expression profiling of endothelial lineage-specificity using vascular gene signatures. EMBO – Advanced Analysis and Informatics on Microarray Data, Cambridge, UK (June 2006). 6. Shin J. Comparison of two microarray-based mRNA profiling technologies for identification of novel endothelial lineage-specific genes. 2nd European Expression Array Systems meeting, Barcelona, Spain (April 2006). 7. Shin J (Young Investigator Award). Lymphatic re-programming of blood vascular endothelium by Kaposi’s sarcoma-associated herpesvirus. Gene Signature Symposium, Zurich, Switzerland (March 2005). POSTER PRESENTATIONS 1. Shin J. Comprehensive identification of endothelial lineage-specific gene signatures and establishment of a low-density microvascular differentiation assay. Gordon Research Conference – Molecular Mechanisms in Lymphatic Function and Disease. Les Diablerets, Switzerland (September 2006). 185 2. Shin J. Comparison of two microarray-based mRNA profiling technologies for identification of novel endothelial lineage-specific genes. Conference for Tumor- Host Interaction and Angiogenesis. Ascona, Switzerland (October 2005). 3. Shin J. Prox1 promotes lineage-specific expression of FGF receptor-3 in lymphatic endothelium: A role for FGF signaling in lymphangiogenesis. 5th International Symposium on the Biology of Endothelial Cell. Dresden, Germany (September 2005). 4. Shin J. Prox1 promotes lineage-specific expression of FGF receptor-3 in lymphatic endothelium: A role for FGF signaling in lymphangiogenesis. 5th ESH Euroconference on Angiogenesis. Sitges, Spain (May 2005). PUBLICATIONS 1. Shin J, Huggenberger R, Detmar M. Transcriptional profiling of VEGF-A and VEGF-C target genes in lymphatic endothelium reveals endocan as a novel mediator of lymphangiogenesis. Manuscript in preparation. 2. Shin J, Detmar M. Lymphatic-specific expression of dipeptidyl peptidase IV and its dual role in lymphatic endothelial function. Manuscript in preparation. 3. Shin J, Fuchs T, Gan W, Kunstfeld R, Mrowietz U, Detmar M. Quantification of vascular lineage-specific differentiation and molecular characterization of in vivo (lymph)angiogenesis by a novel low-density microvascular differentiation array. Manuscript in preparation. 4. Dadras SS, Skrzypek A, Nguyen L, Shin J, Schultz M, Arbiser J, Mihm MC, Detmar M. Prox-1 promotes invasion of kaposiform hemangioendotheliomas. Submitted to Journal of Investigative Dermatology. 5. Detmar M, Shin J. Angiogenesis and lymphangiogenesis in cutaneous cancer. Exp. Dermatology 2007 Oct;16(10): 865-866 6. Shin J, Min M, Larrieu-Lahargue F, Canron X, Nguyen L, Kunstfeld R, Henderson JE, Bikfalvi A, Detmar M, Hong YK. Prox1 promotes lineage-specific expression of FGF receptor-3 in lymphatic endothelium: A role for FGF signaling in lymphangiogenesis. Mol Biol Cell. 2006 Feb;17(2):576-84. 7. Hong YK, Shin J, Detmar M. The development of the lymphatic vascular system- a mystery unravels Dev Dyn. 2004 Nov;231(3):462-73. 8. Hong YK, Foreman K, Shin J, Hirakawa J, Curry CL, Sage DR, Libermann T, Dezube BJ, Fingeroth JD, Detmar M. Lymphatic re-programming of blood vascular endothelium by Kaposi’s sarcoma-associated herpesvirus Nature Genetics. 2004 Jul;36(7):683-5. 186 9 ACKNOWLEDGEMENTS The work presented in this thesis would not have been possible without the support of the following people. First of all, I would like to express my genuine appreciation to my supervisor Prof. Dr. Michael Detmar, who gave me the great opportunity to independently work on numerous exciting projects. Throughout my thesis, Michael provided encouragement, sound advice, good teaching, and positive outlook. I am very much grateful for his generous support during my thesis. I would also like to thank Prof. Dr. Dario Neri for kindly accepting to be co-referent of my thesis and also for his positive motivation during my thesis. I am grateful to Reto Huggenberger for the friendly collaboration and for his exceptional contributions to the ‘endocan’ project during his diploma work. For the ‘LD-MDA’ project, I would like to acknowledge Thomas Fuchs for his brilliant statistical analyses and Dr. Weiniu Gan for the fruitful discussions and his significant contributions to the analysis of ‘core’ signature genes. I would also like to thank Dr. Rainer Kunstfeld (University of Vienna) and Dr. Ulrich Mrowietz (University of Kiel) for their help in obtaining clinical samples and for the productive discussions. I would like to acknowledge Dr. Andreas Bikfalvi, Dr. Frederic Larrieu-Lahargue, Dr. Xavier Canron (from University Bordeaux) for their significant contributions to the ‘FGFR-3’ project. I am also grateful to Jana Zielinski, Cornelius Fischer, Michael Min and Lynh Nguyen for their outstanding technical supports. In addition, I would like to thank all members of Detmar’s group who contributed to the friendly atmosphere and provided numerous unforgettable moments. I am particularly grateful to Nadja Tobler for the ‘thought-provoking’ discussions and her moral support. Special thanks also goes to Leah Cueni for translating my summary. 187 Furthermore, I would like to thank Eveline Trachsel for helping me to assimilate into the Swiss culture and guiding me through my Ph.D. process. Finally, I would like to express my up-most gratitude to Evelyn and to my parents for their unconditional love and support. To them I dedicate this thesis. 188